• An anonymous Microsoft engineer has claimed that managers working on a major oil-extraction project called Tengizchevroil, or TCO, proposed using tech to surveil staff working in their oil fields.
• Writing in Logic Magazine, the engineer recounted how TCO managers wanted to use artificial intelligence, machine learning, and data from GPS trackers to analyze CCTV video footage to track workers' behavior.
• TCO is a joint venture among Chevron; ExxonMobil; the Kazakh state-owned oil firm KazMunayGas; and LukArko, a subsidiary of the Russian oil firm Lukoil. The anonymous engineer was sent to Kazakhstan because of a preexisting arrangement between Microsoft and Chevron.
• Microsoft did not immediately respond to Business Insider's request for comment.
• Visit Business Insider's homepage for more stories.

An anonymous Microsoft engineer has written what appears to be an explosive account of their experiences with a major multinational oil project called Tengizchevroil.

In an essay published in Logic Magazine, the engineer — writing under the pseudonym "Zero Cool"— described how managers overseeing the project wanted Microsoft's help to surveil workers in their oil fields.

The anonymous account doesn't give a timeline for the events described. The writer is described by Logic Magazine as "a software engineer working at Microsoft."

Microsoft, ExxonMobil, and Chevron have not responded to the account and did not immediately respond to requests for comment from Business Insider. The report comes as Silicon Valley workers step up action against their employers for working with major oil firms and other companies that affect climate change.

The project, known as TCO, is a joint venture among the oil giants Chevron and ExxonMobil as well as the Kazakh state-owned oil firm KazMunayGas and LukArko, a subsidiary of the Russian oil firm Lukoil. As such, many of the venture's oil fields are in Kazakhstan, and many of its oil-field workers are local Kazakhs.

The anonymous engineer recounted going to Kazakhstan because of a preexisting arrangement between Microsoft and Chevron to help TCO oil fields adopt Microsoft technology.

The engineer said TCO managers proposed using artificial intelligence and machine learning to track staff.

The engineer wrote: "This is what our Chevron partners were most keen to discuss: how to better surveil their workers."

The person added: "They proposed using AI/ML to analyze the video streams from existing CCTV cameras to monitor workers throughout the oil field. In particular, they wanted to implement computer vision algorithms that could detect suspicious activity and then identify the worker engaging in that activity."

TCO managers said they didn't trust workers

This reportedly wasn't the only surveillance method managers proposed, either.

"The TCO managers also talked about using the data from the GPS trackers that were installed on all of the trucks used to transport equipment to the oil sites," the engineer continued. "They told us that the workers were not trustworthy. Drivers would purportedly steal equipment to sell in the black market."

The idea was to build a machine-learning model that could identify any abnormal driving activity. But such a model would also let managers track workers' productivity and let them monitor bathroom breaks and what routes they took.

One unnamed manager reportedly said it "would be nice" to "know where they are and what they are doing, if they are doing anything at all."

Microsoft signed a seven-year deal with Chevron in 2017, which was billed as an effort to "digitize" oil fields.

The engineer's allegations compound what is already a widely criticized relationship between Microsoft and the oil industry.

In September, Microsoft employees signed an open letter critical of the tech giant's work with oil firms, and the Gizmodo reporter Brian Merchant accused Microsoft and other tech giants of "automating the climate crisis" in a stinging op-ed article.

SEE ALSO: Big Oil claims it's doing its part to combat climate change. A new study finds it's not even close.

Join the conversation about this story »

NOW WATCH: How autopilot on an airplane works

• Nines Radiology is focused on using technology, including artificial intelligence, to help radiologists.
• Nines was founded by David Stavens, a self-driving-car pioneer, and Dr. Alexander Kagen, a New York City radiologist.
• The company has created a product that is supposed to help radiologists prioritize their time for patients most in need of treatment and diagnose them faster. 
• The product, called the Emergent Neuro Suite, is now under FDA review.
• The startup just raised $16.5 million from investors including the venture firms Accel and 8VC. 
• Click here for more BI Prime stories.

It was a nontraditional marriage of tech and medicine. But when Dr. Alexander Kagen and David Stavens first met, they believed their vision for healthcare could save lives. 

In 2017, they founded Nines Radiology, a startup focused on advanced technology applications for radiologists— doctors who make diagnoses based on medical imaging. The startup just raised a Series A funding round, taking in $16.5 million from investors including the venture firms Accel and 8VC. 

Part of the company is a teleradiology service, staffed by specialists from institutions like Mount Sinai Health System. Teleradiology allows radiologists to look at X-rays and other imaging remotely and provide diagnoses for patients.

The other part of the company works on artificial-intelligence and machine-learning applications. Nines created the Emergent Neuro Suite, a triage tool that is supposed to help radiologists prioritize their time for patients most in need of treatment. The goal is to improve patient and doctor experiences with more efficient and accurate diagnoses, and the product is being reviewed by the Food and Drug Administration.

"The idea is to allow radiologists to get to the sickest patients first, which is the best we can do in terms of quality and safety," Kagen, the chief medical officer of Nines, told Business Insider. 

Kagen oversees the radiology team in New York, while Stavens oversees the engineering team in San Francisco. 

Teleradiology is estimated to be an $11.5 billion industry by 2026, with the fastest area of growth coming from CT scans, according to a report by Grand View Research.

According to a Business Insider Intelligence report, the number of FDA-approved AI solutions for radiology is growing faster than other applications of AI in imaging. AI-generated healthcare savings could be more than $150 billion by 2025, as AI can improve outcomes and reduce treatment costs, according to the report.

How an AI expert and radiologist formed their partnership 

In 2016, Stavens and Kagen met through a mutual friend and hit it off. Both saw the benefits of marrying AI and radiology to provide health benefits for patients and help radiologists manage their workload.

While each come from very different backgrounds, they saw the technological need in radiology. 

Kagen, the site chair of radiology at Mount Sinai West and Mount Sinai St. Luke's hospitals in New York City, has been interested in AI for some time. In 2014, while doing research for Parkinson's disease, he saw how a simple algorithm could accurately predict if someone had Parkinson's. 

"I saw it as the future of medicine," Kagen said. 

Stavens said Kagen understood the need for AI in radiology. 

"He was one of the early adopters of the mind-set," Stavens, the CEO of Nines, told Business Insider. "He wanted to be an agent of change and help design the technology." 

Stavens brought to the table a wealth of knowledge of AI. He cofounded Stanford University's self-driving-car team, which was later acquired by Google and is now the foundation for Waymo. He then cofounded and was CEO of Udacity, which used AI to help build an online learning platform. 

Stavens said he couldn't have predicted he'd be in the radiology or diagnostics space, but as he says, his career has taken unpredictable turns for a reason. 

"The driving force of my career has been applying advanced technology to solve problems that affect humanity," Stavens said. 

In healthcare, Stavens said data driven analysis was being used more to improve medical diagnosis. "I see the need for technology in healthcare. Its time has come." 

Why radiology is ready for AI

Right now, radiologists must look at thousands of imaging scans for patients and make accurate diagnoses. The sheer volume of scans is becoming overwhelming for radiologists to analyze, Kagen said. 

In order to help radiologists with the volume of data help patients see their physicians faster, Stavens and Kagen wanted to create a product that would be able to detect the sicker patients in the queue of scans. An alert would then be sent to the radiologist to evaluate the scan faster and help the patient. 

"Maybe patient 10 in the queue is sicker than patient four," Kagen said. "It would be great if there was a way where we could define patient 10 and flag the radiologist."

For Kagen, an important aspect of Nines is that the radiologists at Mount Sinai can communicate with the engineers in Silicon Valley and tell the engineers specifically which technological improvements the product needs. 

For the cofounders, giving radiologists technological input is vital for the engineers to create the strongest product. 

"We'll implement this software with teleradiology services," Kagen said. "Radiologist can play an active and prominent role working with engineers, allowing them to be the best superdoc they can be."

Join the conversation about this story »

NOW WATCH: A 45-year-long study discovered trends in successful hyper-intelligent children

• Researchers successfully built a quadruped robot that's capable of climbing ladders.
• The robot was created by a team at Tokyo Metropolitan University and debuted earlier this month, IEEE Spectrum reported.
• It uses a 3D camera, touch and force sensors on its claws. A neural network — learning software that allows the robot to execute complex functions — enables it to navigate ladders.
• Visit Business Insider's homepage for more stories.

Four-legged creatures, both living and robotic, have historically had a hard time climbing ladders — until now.

Researchers at Tokyo Metropolitan University have created a quadruped robot capable of climbing a vertical ladder without assistance, IEEE Spectrum reported. 

The robot debuted at the International Conference on Intelligent Robots and Systems earlier this month.

The achievement represents the latest physical feat by a quadruped robot that approaches the capabilities of living animals. Robot dogs made by other research teams have been trained to open doors, withstand destabilizing blows, and run across uneven ground.

While there's no single intended purpose researchers have in mind for quadruped robots, government agencies are already expressing interest — earlier this year, Massachusetts police began working with "Spot," a robot dog designed by Boston Dynamics.

Take a look at how Tokyo Metropolitan University's robot climbs ladders, bringing robots one rung closer to world domination:

SEE ALSO: From 'Jeopardy' to poker to reading comprehension, robots have managed to beat humans in all of these contests in the past decade

The robot weighs in at just over 15 pounds, and is equipped with 3D cameras on its head and touch and force sensors on its claws.

The robot has 23 degrees of freedom — specific aspects of its parts can move: five degrees of freedom in each leg, two for the dual laser rangefinder sensors, and one for the head.

The robot uses a neural network to automatically teach itself to balance as it climbs the ladder.

Once at the top of the ladder, the robot uses its rear claws to grip the top rung and shift its weight forward.

The robot was specifically programmed to climb the ladder used in the demonstration video, but researchers plan to teach it to climb a ladder of any dimensions.

While living quadruped animals can climb ladders at an angle, they struggle with vertical ladders, according to IEE Spectrum — but this robot is capable of tackling them.

Watch the full video of the ladder-climbing robot demonstration:

Youtube Embed:
//www.youtube.com/embed/k-SdJ1clNMQ
Width: 560px
Height: 315px

• Lee Se-dol, a former Go champion, has announced his retirement, declaring that artificial intelligence has created an opponent that "cannot be defeated."
• Go, a strategic board game that originated in China more than 2,500 years ago, is considered one of the most complex games in the world.
• In 2016, Lee lost a set of matches against AlphaGo, an artificial intelligence developed by DeepMind, a company owned by Google.
• DeepMind's programs have mastered several games and defeated top human competitors, bringing new attention to the rapid growth in AI technology.
• Visit Business Insider's homepage for more stories.

The dominance of artificial intelligence in competitive strategy games has led one of South Korea's top Go players to retire.

Lee Se-dol, 36, told the Yonhap News Agency that he would no longer play professionally, because AI is impossible to overcome.

"With the debut of AI in Go games, I've realized that I'm not at the top even if I become the No. 1 through frantic efforts," Lee said. "Even if I become the No. 1, there is an entity that cannot be defeated."

In 2016, Lee, who was considered the best Go player in the world, lost a set of matches 4-1 against AlphaGo, an AI system developed by the Google-owned company DeepMind.

Dating back more than 2,500 years to China, Go is considered one of the most complex board games in the world, and AlphaGo's mastery brought new attention to the rapid advancement of AI technology.

DeepMind has further improved its program since AlphaGo's match against Lee too. The latest iteration, AlphaGo Zero, was able to beat the original AlphaGo in 100 matches in a row.

DeepMind's AI programs improve by analyzing thousands of professional matches, allowing them to anticipate the opponent's moves and counter with the most effective strategies and counterplays.

Lee will have one last match against an AI opponent before he officially retires in December: He'll be up against HanDol, an AI developed in South Korea that's already defeated the country's top five players. Lee will be given a slight handicap against HanDol, but he still isn't confident about how the match will go.

"Even with a two-stone advantage, I feel like I will lose the first game to HanDol," Lee told Yonhap.

DeepMind's AI programs have mastered some of the most popular strategy games, from classic board games like chess and Go to competitive video games. Last month, DeepMind's "StarCraft II" AI program, AlphaStar, reached the rank of grandmaster, making it better than 99.8% of all human players. AlphaStar has also defeated some of the world's best esports competitors in head-to-head matches across different games.

While AI's superiority over human players seems to be an inevitability, these programs have also helped to push the boundaries of these strategy games, and matches against the world's top players evoke the highest possible level of play.

SEE ALSO: Google's DeepMind is better than 99.8% of human players of 'Starcraft II,' putting it on the long road towards solving real-world problems

Join the conversation about this story »

NOW WATCH: Apple just released iOS 13.2 with 60 new emoji and emoji variations. Here's how everyday people submit their own emoji.

• The Department of Defense released its artificial intelligence strategy in February, writing that it is working toward "the creation of a force fit for our time." 
• To accelerate the pace of innovation, the Pentagon is increasingly relying on Other Transactions Authority (OTA) contracts, creating more prototyping, research and development opportunities.
• "What they're allowing us to do through the OTA process is they're creating an environment where it's okay to innovate," a Raytheon senior solutions architect told Business Insider.
• OTA prototyping jumped from $1.4 billion to $3.7 billion, with the number of awards rising from 248 to 618, between 2016 and 2018, the Government Accountability Office revealed in a recent report.

The Department of Defense is giving its industry partners room to make mistakes, explore, and innovate in the field of artificial intelligence through the use of non-traditional contracts, an industry expert explained to Business Insider in a recent interview.

AI is not like the atomic bomb, but it is a revolutionary enabler, something like electricity, and it is being intensely explored by both the US and its great power rivals.

The US military is striving to make advancements in AI for "the creation of a force fit for our time," the DoD Artificial Intelligence Strategy released in February says, adding, "We cannot succeed alone."

One of the ways the Pentagon is bypassing the bureaucracy that has traditionally been a hindrance to rapid innovation is Other Transactions Authority (OTA) contracts.

The defense department is increasingly using this type of contract, a special agreement for department prototyping, research, and development projects, to make forward progress in new areas.

Between 2016 and 2018, OTA prototyping jumped from $1.4 billion to $3.7 billion, a recent Government Accountability Office (GAO) report revealed. The number of OTA prototyping contracts increased nearly 150 percent from 248 to 618.

"What they're allowing us to do through the OTA process is they're creating an environment where it's okay to innovate," Corey Hendricks, senior solutions architect at Raytheon, recently explained to Business Insider.

"It's okay for us to bring them the latest technology advancements that haven't gone through the rigor of extensive testing," he added, explaining that DoD has "allowed us to bring them a prototype where there might be a couple of wires hanging out, if you will, but they can see a roadmap where this is a viable product."

Hendricks characterized the OTA approach as a "fantastic change" from the traditional contracting process, which put a lot more restrictions on innovation.

This contracting option, which has been around for decades but has recently been expanded, gives industry more freedom to address obstacles to AI development, such as computing and communications.

Raytheon, a leading defense contractor, is currently looking into the use of AI and machine learning for US military training, specifically simulated training environments and real-time analysis.

"It also brought a large corporation like Raytheon closer to small businesses," he further explained. "It really does cultivate that culture of innovation."

The GAO revealed in its recent report that 88 percent of the OTA contracts awarded since 2016 went to companies with which the Pentagon has not traditionally done business with, including many small businesses.

While more extensive use of OTAs is an improvement, it remains to be seen how DoD intends to move past the preliminary innovation and prototyping stages to the mass production of new technologies.

Join the conversation about this story »

NOW WATCH: We can thank the US military for the smelliest weapon in the world

• The rise of artificial intelligence is giving companies capabilities they've never had before to automate job tasks, recruit top talent, and better know their customers, among other applications. But adopting the advanced tech is not an easy shift, and many efforts still fail.
• Business Insider is exploring how companies can successfully implement the tech without succumbing to the roadblocks that stymie so many efforts.
• These reports can help prepare executives as they ready themselves to lead the AI-efforts, as well as provide guidance for getting the enterprise onboard. 
• Business Insider regularly interviews executives about their company's AI efforts. You can read them all by subscribing to BI Prime.

Companies across corporate America are harnessing artificial intelligence and machine learning to, among other things, automate the more mundane-aspects of jobs, help to recruit top talent, and know more about customers to tailor promotional offers or product recommendations. 

But the move to adopt the advanced technology comes with significant challenges and often requires a major cultural shift. It's one reason why many AI-based projects still fail.

Business Insider is exploring how companies can successfully implement the tech without succumbing to the roadblocks that stymie so many efforts.

Why culture is more of a limitation than tech in the push to AI: CarMax's CIO says culture is more of an impediment than technology for organizations pursuing a digital overhaul. Here's how he managed the shift at the nation's largest used car retailer.

How Frito-Lay is trying to overcome AI hurdles: Frito-Lay's transformation chief outlines the biggest challenges facing the digital overhaul effort at $189 billion PepsiCo's most profitable division

Using AI to predict fast food orders:All the ways McDonald's is using AI to learn what you'll order — sometimes even before you know what you'll want

Why tech companies struggle with AI culture:IBM's global chief data officer explains why changing culture to support AI is harder in a traditional tech company — and one way it's easier

How to kickstart your AI effort successfully:Here are the 3 steps companies that are just pivoting to AI should take to guarantee the process starts on a solid footing

Using AI to find better talent:CareerBuilder used artificial intelligence to figure out veterinary technicians could make good prison guards. It's just one way the hiring platform is employing the technology to revolutionize HR.

Learning from an AI leader:I spent a day at IBM's mysterious research hub north of NYC, where I met some of the top AI leaders in the country. Here are 4 takeaways on where they think the tech is headed.

How Walmart evaluates AI projects:Walmart has 1,500 data scientists and is hiring more amid a push to adopt artificial intelligence. The retailer's chief data officer recently shared the 3 questions that guide all its AI projects.

Regulating AI is still a key challenge:There are no laws regulating the use of AI in the hiring process, and it's setting back how companies recruit. Here are the people trying to change that.

What to weigh before pouring resources into AI:Accenture's head of artificial intelligence shares the 4-step plan every company should consider before investing in AI

Using employees as AI testers:Walmart has cracked the code for merging AI rollouts with employee feedback to produce buzzy (and cost-saving) new tech

Using AI to improve retail operations:'It's the art of the possible': How Walmart and Target are harnessing AI to rocket past the competition

SEE ALSO: THE CHANGING C-SUITE: What the rise of information, data, and tech chiefs says about the future of leadership in America's top companies

Join the conversation about this story »

NOW WATCH: Taylor Swift is the world's highest-paid celebrity. Here's how she makes and spends her $360 million.

• Google has added the cofounder of its DeepMindartificial intelligence unit to its own internal AI team.
• Mustafa Suleyman, who cofounded DeepMind with Demis Hassabis and Shane Legg, took a temporary leave from the firm earlier this year.
• His move to Google ends speculation that he planned to quit both DeepMind and Google altogether.
• Google acquired DeepMind for 400 million pounds in 2014, and the firm remains headquartered in London and a research-oriented entity.
• Visit Business Insider's homepage for more stories.

Google has added the cofounder of its UK-based DeepMind artificial intelligence research unit to its own AI team.

Mustafa Suleyman confirmed the news on Twitter on Thursday, saying that he would join Google policy chief Kent Walker and Jeff Dean, the senior vice president of Google AI, in January.

"After a wonderful decade at DeepMind, I'm very excited to announce that I'll be joining @Kent_Walker, @JeffDean and the fantastic team at Google to work on opportunities & impacts of applied AI technologies," Suleyman said in his tweet. "Can't wait to get going! More in Jan as I start the new job!" 

Suleyman will be working on AI policy issues for Google, a company representative told Business Insider. The representative did not say what Suleyman's title will be.

In tweets, both Walker and Dean welcomed Suleyman to Google.

Meanwhile, in a blog post, DeepMind CEO Demis Hassabis praised Suleyman's work with the unit. Google and parent company Alphabet are increasingly incorporating DeepMind's AI work into their other divisions, thanks in large part to Suleyman's efforts, he said.

"Mustafa played a key role over the past decade helping to get DeepMind off the ground, and launched a series of innovative collaborations with Google to reduce energy consumption in data centres, improve Android battery performance, optimise Google Play, and find ways to improve the lives of patients, nurses and doctors alike," Hassabis wrote. 

Suleyman also helped spearhead the company's controversial health division. Two years ago, that division was found to have improperly obtained access to the health records of 1.6 million UK patients.

After Suleyman announced a temporary leave from DeepMind earlier this year, there was speculation he might quit the company and Google altogether. That move coincided with Google's announcement that it would absorb part of DeepMind's health-focused operations into its own broader healthcare unit, Google Health.

Suleyman founded DeepMind in 2010 with Hassabis and Shane Legg. Google acquired the company for 400 million pounds ($486 million) in 2014.

SEE ALSO: The cofounder of Google's AI company DeepMind hit back at 'speculation' over his leave of absence

 Mustafa Suleyman: The liberal activist who cofounded Google's $486 million artificial intelligence lab

Join the conversation about this story »

NOW WATCH: Inside the US government's top-secret bioweapons lab

• The founder of self-driving truck firm TuSimple, Xiaodi Hou, says that truck drivers being required to work long hours on the road is a "tarnish on the glory of humanity."
• Hou was responding to a question about human truckers who fear that automated trucking might lose them their jobs.
• TuSimple, which is valued at over $1 billion, develops tech to automate long-haul truck journeys without human intervention needed, though its vehicles still have a human as a failsafe. The company straddles both China and the US.
• Speaking to Business Insider, Hou also explained why traditional trucking associations are actively working with, and not against, his firm.
• Visit Business Insider's homepage for more stories.

The founder of self-driving truck unicorn TuSimple says that human truck drivers being required to work long hours on the road is a "tarnish on the glory of humanity."

Xiaodi Hou is the founder, president and chief technology officer of TuSimple, which develops tech for automating long-haul truck journeys without human intervention needed. The self-driving trucks still have a human truck driver and an engineer present at all times, as a failsafe.

The company is based in San Diego, California, and splits its operations with China. The firm's US presence comprises offices in San Diego, California and Tucson, Arizona, from which it oversees its long-haul US operations.

It also boasts offices in Beijing, Shanghai, and Fukuoka, Japan. Hou said he will expand its presence further in the next two to three years – a goal he said stemmed from "the confidence of our general stability."

Founded in 2015, TuSimple has enjoyed huge investor interest. It was valued at just over $1 billion after a $95 million Series D fundraise in February. It went on to add a further $120 million, bringing the round to $215 million. Its backers include global delivery giant UPS, US chipmaker Nvidia, and Chinese tech firm Sina, owner of Weibo.

Though TuSimple currently retrofits existing trucks for pre-existing truck companies, it aims to have a factory-produced self-driving truck on the road by 2023.

TuSimple doesn't think it will completely replace drivers with self-driving trucks

Speaking to Business Insider, Hou was asked how he'd respond to truckers who (rightly or wrongly) fear that their jobs are at risk.

"This is my first time responding to this question in English, but here goes," he said. "To drive a truck for 11 hours per day, without even taking a shower every day, and getting far away from their home, is really a tarnish on the glory of humanity.

"You don't hire chimney sweeps nowadays, and people don't harvest the way they did 500 years ago."

Asked what he'd say to those truckers who value the work, Hou said his company complements – and doesn't compete with – the traditional trucking industry.

"The transition is actually slower than you thought," he explained. "The transition is not like, 'tomorrow, all of sudden, trucks will be autonomously driven.' It's not like that. Autonomous driving is more likely to involve the retrofitting of existing trucks rather than the building of new trucks.

"It'll be a very gradual thing, and truck drivers still need to [drive] trucks. Think about the gap we have [in terms of] truck driver shortage. The average truck driver age is 51. I don't think even [on] a very optimistic view, we can fill up that gap within the time frame it needs to be filled."

According to a July 2019 report by the American Trucking Associations, the US trucking industry was 60,000 drivers short of the number required to meet industry needs in 2018, up nearly 20% from 2017's shortage of 50,700.

"Secondly, we don't actually think self-driving trucks are going to deprive any truck drivers [of their jobs]," Hou continued. "I don't think we're going to be enemies of the truck drivers. We're actually very good friends with the American Trucking Associations.

"We're helping the associations to solve problems that they can't solve alone, by themselves."

SEE ALSO: The self-driving-truck startup that's been quietly moving UPS loads just pulled in $120 million in fresh funding

Join the conversation about this story »

NOW WATCH: Why it's so hard for planes to land on water

• The impact of artificial intelligence on the workforce is a hotly debated subject. In the world of gaming, however, AI is already replacing humans.
• Facebook AI researchers say they built a bot that can exceed human performance in Hanabi, a cooperative game that requires players to interpret the actions and intentions of other players. 
• The achievement signals a jump in the capability of the technology to replicate the cognitive skills of humans, as opposed to the method of training the models through rewards and punishments that was used for other models. 
• Click here for more BI Prime stories.

Opinions are split on the magnitude of the impact that artificial intelligence will have on the workforce. In the gaming world, however, it's already displacing top human players.

Last month, Lee Se-dol — one of the top South Korean players in Go, a complex Chinese board game that dates back 2,500 years — announced he would retire because it became impossible to beat AI-powered bots. This comes after Se-dol lost a series of matches in 2016 to a bot developed by Google-owned AlphaGo.

Now, Hanabi may be next. Facebook researchers have built a bot they say can outperform the most skilled players at the cooperative card game that requires individuals to play off the hands of other participants because they are unable to view their own cards.  

While other models have exceeded at games like Go, chess, and poker, researchers were able to scale those projects through a technique known as "reinforcement learning," a process that uses a series of rewards and punishments to train behavior — similar to how one might approach raising a puppy.

Project leads at the social media giant's research lab say their latest bot signals a significant improvement because it can replicate cognitive skills to put itself in the perspective of other players in the game, a trait known as "theory of mind."

"These techniques that we are developing that can cope with things like theory of mind will play an important part as we extend to other domains like natural language processing," AI researcher Noam Brown told Business Insider.

Employing AI to cooperate, instead of compete, against humans

Facebook AI Research is a group of about 300 employees that is wholly focused on academic studies into the technology. While some of the projects might dovetail off of Facebook's AI goals, Brown and fellow researcher Adam Lerer say they independently choose what problems to tackle.

The group aims to answer "long-term research questions that we think of as barriers to the advancement of artificial intelligence," said Lerer.

"You can't always predict what is going to be the next big thing, so you have to give the freedom to the researchers … even if there's not a five-year plan for turning this into a product" that can be monetized, Brown added.

Hanabi posed a unique challenge because unlike chess, for example, the bot is not going strictly head-to-head against a human. The nature of the game requires players to cooperate with other participants and interpret their intentions based on gameplay to win.

A player is unable to view their hand — just the hands of the other participants — and only narrow hints are provided throughout gameplay. So one player may seek to gain an advantage by suggesting one hidden card is a certain color because he or she wants that color played next. 

Hanabi was proposed as a key challenge for AI by researchers at Google's DeepMind for that reason. 

The model had to reason "about the intent and the beliefs that other people have that lead them to take the actions that they do." Lerer said. The team had to build a "conversational agent that can reason about what a person is saying without having to explicitly tell them everything."

The project is so promising because it signals that AI could eventually be able to cooperate alongside humans to tackle complex challenge, as opposed to just replacing them in tasks like monitoring shelves in Walmart for restocking.

SEE ALSO: Accenture's head of artificial intelligence shares the 4-step plan every company should consider before investing in AI

Join the conversation about this story »

NOW WATCH: Taylor Swift is the world's highest-paid celebrity. Here's how she makes and spends her $360 million.

• Jobs in banking and the financial services industries continued to be among the most popular occupations of 2019.
• Despite their popularity, a report predicts that 1.3 million bank workers will lose their jobs or be reassigned due to automation.
• Banks have already begun investing in artificial intelligence, and recognize the technology will displace workers.
• Visit BusinessInsider.com for more stories.

Jobs in banking are some of the most sought after for job seekers — but plenty of roles may not be around much longer. 

Algorithms that model prices or build portfolios could wipe out 6 million high-paying jobs in finance, according to Cornell University professor Marcos Lopez de Prado. 

Lopez de Prado told the US House Committee on Financial Services that AI might not replace jobs entirely, but current finance employees aren't trained to work alongside new technology, Bloomberg reported. 

Lopez de Prado's statement aligns with a 2019 report that revealed 1.3 million US finance jobs — particularly customer-service reps, financial managers, and compliance and loan officers — could disappear by 2030, according to a British insights firm IHS Markit. Brookings, too, recently found white-collar employees in tech and finance are more susceptible to AI job loss than social workers, teachers, or cooks.

How AI is set to disrupt the finance industry

Despite a year of scandals that entangled many of the country's largest banks, the desire to work at these companies remains high, according to a report by LinkedIn. Some of the more high-profile scandals include Deutsche Bank's alleged involvement in a global money-laundering scheme and accusations against Wells Fargo's auto-loan and mortgage practices.

Nonetheless, Bank of America, Goldman Sachs, Citigroup, Wells Fargo, and JPMorgan Chase remain five of the most popular places to work in 2019. LinkedIn attributes the popularity to banks offering increasingly tech-focused jobs that attract talented software engineers and developers out of college.

"The reality is that if somebody wants to learn finance and strategy, these banks are still the places to be trained and developed," Heather Hammond, co-head of the global banking and markets practice at Russell Reynolds Associates, told LinkedIn.

Jobs in banking as a whole are some of the most expensive in the country. Starting analysts make $91,000 in base pay, while managing directors can earn almost $1 million after bonuses. In fact, the industry could add a whopping $512 billion in global revenue by 2020 with the use of intelligent automation, according to a 2018 report from Capgemini.

Major banks have already begun implementing AI

While the use of AI remains sparse, and the technology is still basic, a boost in revenue will increase the adoption of automation, Business Insider analyst Lea Nonninger reports.

Unfortunately for job seekers, banks' investment into automation is well under way. In fact, a detailed 2018 report from Business Insider Intelligence noted that banks are already using AI to mimic bank employees, automate processes, and preempt problems. JPMorgan is cleaning thousands of databases to make room for machine learning tech. Citi president Jamie Forese said in 2018 that robots could replace as many as 10,000 human jobs within five years.

Laura Barrowman, chief technology officer at the Swiss investment bank Credit Suisse, revealed the company is already retraining employees whose jobs have been displaced by AI: "Globally, if you look at cyber skills, I think there is a deficit," Barrowman told Business Insider's panel at the World Economic Forum earlier this year. "There is such a shortage of skills, and you need people who have that capability."

SEE ALSO: Here's exactly what it takes to get a job as a banker at Goldman Sachs, according to Wall Street recruiters, current and former employees, and the head of HR

Join the conversation about this story »

NOW WATCH: Taylor Swift is the world's highest-paid celebrity. Here's how she makes and spends her $360 million.

• North American banks could save $70 billion by 2025 by using technology such as automation and artificial intelligence to trim workers and boost productivity, according to a Thursday study from Accenture. 
• Savings across the entire financial services industry, which includes banking, insurance, and capital markets, could reach $140 billion in the same time frame, according to the report. 
• "Automation and augmentation can liberate hours that can then be refocused on human strengths and high-value work such as innovation, relationships and customer experience," according to the report.
• Read more on Business Insider.

Banks could save billions of dollars by using technology such as automation and artificial intelligence, according to a study released Thursday by Accenture. 

By using technology to automate jobs or help employees at work, North American banks could save more than $70 billion by 2025, the study showed. 

Across the entire financial services industry, which includes banking, insurance, and capital markets, savings could be between $87 billion and $140 billion in the same time frame. 

"Massive social and technology change is creating a range of threats and challenges," for financial services firms, Accenture wrote. But, it also creates "the opportunity to unlock significant value through a new workforce, new ways of working, and new job roles." 

The study estimates that by 2025, 7% to 10% of tasks will be automated, boosting cost and productivity savings.

"Automation and augmentation can liberate hours that can then be refocused on human strengths and high-value work such as innovation, relationships and customer experience," according to the report. These are things that also generate "substantial value" for organizations, Accenture said. 

More than half of tasks currently performed by loan officers, personal financial advisers, tellers, and customer service representatives could be either automated or augmented by technology by 2025, the study found. 

A number of banks such as Citigroup, Capital One, and JPMorgan Chase are already using technology like artificial intelligence to aid workers, or have automated parts of tasks to eliminate jobs. 

Studies have estimated that as many as 1.3 finance jobs in the US could disappear by 2030 as banks look for ways to cut costs.

Join the conversation about this story »

NOW WATCH: A big-money investor in juggernauts like Facebook and Netflix breaks down the '3rd wave' firms that are leading the next round of tech disruption

• Grimes, the Canadian singer, performed Thursday night at the Game Awards, the Oscars of the video game industry, which celebrated the year's best games.
• A notable face in the audience was Elon Musk, the Tesla CEO who Grimes has been dating since 2018.
• Here's everything you need to know about Grimes, who has been alongside Musk as he faced off against the SEC over his infamous "funding secured" tweet.
• Visit Business Insider's homepage for more stories.

At the Met Gala in 2018, a surprising new couple showed up on the red carpet: billionaire tech CEO Elon Musk and Canadian musician and producer Grimes.

While Musk has long been known to date successful and high-profile women, the two made a seemingly unlikely pairing. Shortly before they walked the red carpet together, Page Six announced their relationship and explained how they met — over Twitter, thanks to a shared sense of humor and a fascination with artificial intelligence.

Since they made their relationship public in May 2018, the couple has continued to make headlines: Grimes for publicly defending Musk and speaking out about Tesla, and Musk for tweeting that he wants to take Tesla private, sparking an SEC investigation. But shortly after Musk's run-in with the SEC, Grimes and Musk unfollowed each other on social media, igniting rumors that the pair had broken up. 

However, the couple appeared to have reconnected soon after, and have been spotted out together. Grimes was in the car when Musk was spotted driving his new Cybertruck prototype around Los Angeles, and Musk made an appearance at the 2019 Game Awards to watch Grimes' performance at the event. 

For those who may be wondering who Grimes is and how she and Musk ended up together, here's what you need to know about the Canadian singer and producer.

SEE ALSO: Everything you need to know about e-girls and e-boys, teen gamers who have emerged as the antithesis of Instagram influencers

Grimes — born as Claire Boucher — grew up in Vancouver, British Columbia. She attended a school that specialized in creative arts, but didn't focus on music until she started attending McGill University in Montreal.

Source: The Guardian, Fader

A friend persuaded Grimes to sing backing vocals for his band, and she found it incredibly easy to hit all the right notes. She had another friend show her how to use GarageBand and started recording music.

Source: The Guardian

Grimes has been described as an electronic-pop artist, whose music is "dark and ethereal, catchy and strange." In a 2019 profile, the Wall Street Journal described her music as "the kind of music you imagine a group of vampires would listen to if this group of vampires also happened to be on a cheerleading squad."

Source: Wall Street Journal

In 2010, Grimes released a cassette-only album called "Geidi Primes." She released her second album, "Halfaxa," later that year and subsequently went on tour with the Swedish singer Lykke Li. Eventually, she dropped out of McGill to focus on music.

Source: The Guardian, Fader

In 2012, Grimes signed to the British indie label 4AD and released "Visions," which would become a breakout success. Two years later, Pitchfork named "Oblivion" the best song of the decade so far.

Source: Pitchfork, The Guardian

Grimes signed with Jay-Z's management company, Roc Nation, in 2013.

Source: Fader

As she worked on the follow-up album to her widely acclaimed "Visions" album, Grimes considered a move to Billboard-charting pop music. She even wrote a song in 2014 for Rihanna, who ended up not putting the song on her album. Grimes released the song herself, but fans were upset she was "pandering to the radio."

Source: New York Times

Grimes released her fourth studio album, "Art Angels," in the fall of 2015. The single of the album, "Flesh Without Blood," features a character she created named Rococo Basilisk who is "doomed to be eternally tortured by an artificial intelligence, but she's also kind of like Marie Antoinette," she told Fuse.

Source: Business Insider, Fuse

Beyond singing, Grimes is a producer, and she's been vocal about how the music industry and media treat female artists. "The thing that I hate about the music industry is all of a sudden it's like, 'Grimes is a female musician' and 'Grimes has a girly voice,'" she told the Fader. "It's like, yeah, but I'm a producer, and I spend all day looking at f---ing graphs and EQs and doing really technical work."

Source: Fader

Grimes is also an avid gamer, and she has streamed herself playing the fantasy role-playing game "Bloodborne" on Twitch, the video-game-streaming platform. Her handle, "@Ocarina_of_Grimes," is a reference to the video game "Legend of Zelda: Ocarina of Time," released in 1998.

Check out her Twitch channel »

Source: Noisey

In May 2018, Grimes attended the Met Gala with Elon Musk, the CEO of Tesla and SpaceX. At the time, reports said they had been "quietly dating" for the past few weeks. Musk later told the Wall Street Journal that he loves Grimes for her "wild fae artistic creativity and hyper intense work ethic."

Source: Page Six, Wall Street Journal

Grimes and Musk met on Twitter. Musk was planning to make a joke about artificial intelligence — specifically, about the Rococo Basilisk character in her "Flesh Without Blood" video — and discovered she had beaten him to the punch.

Source: Page Six

Since then, Grimes has taken to Twitter several times to defend Musk and Tesla. In since-deleted tweets, Grimes said Musk has never tried to stop Tesla workers from unionizing, and claims to have encouraged a union vote among Tesla employees. In a 2019 interview, Grimes said she was "simply unprepared" for how much attention her tweets would get.

Source: Business Insider, Wall Street Journal

Grimes contributed her talents to a song on Janelle Monae's album, "Dirty Computer," released in April 2018. After initially teasing an album of her own for 2018, she said on Instagram that she wouldn't be releasing new music "any time soon" and alluded to a rift between her and her label, 4AD.

Source: Spin

In July 2018, Grimes wrote on Twitter that she and rapper Azealia Banks were collaborating on a song. A month later, Banks flew to Los Angeles to work on music with Grimes at one of Musk's properties, which was the beginning of a tumultuous story involving Banks, Grimes, and Musk.

Source: Business Insider

The day before Banks arrived in LA, Musk posted the now-infamous "funding secured" tweet about having enough money to take Tesla private. Banks said she overheard Musk "scrounging for investors" while at his house, and that Grimes and Musk essentially went into hiding as Tesla sought funding. Banks compared her stay to "a real life episode of 'Get Out,'" and said the couple got her to stay with the promise of collaborating on music.

Source: Business Insider

The relationship between Banks and Grimes devolved from there. Banks later shared with Business Insider a series of messages between the two singers, where Grimes states: "he got into weed cuz of me and he's super entertained by 420 so when he decided to take the stock private he calculated it was worth 419$ so he rounded up to 420 for a laugh and now the sec is investigating him for fraud."

Source: Business Insider

Musk deleted his Instagram and unfollowed Grimes on Twitter in August 2018, prompting rumors that the couple may have broken up.

Source: Business Insider

In September 2018, the Securities and Exchange Commission sued Musk on charges that he made "false and misleading statements" about taking Tesla private at $420 per share. As part of the filing, Musk said he chose that price-point because he thought his girlfriend, Grimes, "would find it funny, which admittedly is not a great reason to pick a price." Musk later settled with the SEC.

Source: Business Insider

It seems that Grimes and Musk reconnected in October 2018: the couple was spotted at a pumpkin patch with Musk's five sons. It was the first time they were spotted in public together since the SEC investigation. Musk also re-followed Grimes on Twitter after a period of flip-flopping on his follow.

Source: Business Insider

Grimes kept a low profile following the "funding secured" scandal. In January 2019, she tweeted she was "randomly" in China on the same day Musk was there to launch Tesla's new factory project in Shanghai.

Source: Business Insider

Grimes returned to the public eye with a profile in the Wall Street Journal's magazine in March 2019. Grimes — who said she now prefers to go by the name "c" instead of her birth name, Claire — referred to Musk as "a super-interesting g------ person."

Source: Wall Street Journal

In November, Musk unveiled Tesla's new battery-powered Cybertruck. The truck was unveiled by a mysterious "cybergirl" hologram, who many have speculated was actually Grimes (although that hasn't been confirmed). The hologram sported a leg tattoo that appears to match one Grimes has.

Source: Business Insider

Grimes' possible appearance as the Tesla "cybergirl" would be fitting, given the e-girl aesthetic that the alt-pop singer portrays. The "e-girl," popularized by teens in 2019, has become a term for the modern-day scene teens who sport a grungy vibe, love video games, and shirk the mainstream, manicured Instagram aesthetic.

Source: Business Insider

A month after the Cybertruck was unveiled, Musk was spotted driving his company's new prototype vehicle around Los Angeles. Grimes was spotted in the truck alongside Musk, who was seen in a video mowing down a traffic sign.

Source: The Cut, Business Insider

Grimes performed Thursday at the 2019 Game Awards, which celebrates the best of the video game industry. Grimes debuted a new song, called "4ÆM," and it was revealed her music — and voice, for a character named Lizzy Wizzy — would be featured in a not-yet-released video game called "Cyberpunk 2077." Musk made a surprising cameo in the audience to watch Grimes' performance, and even gave a short standing ovation at the end.

Source: Business Insider

The song she debuted at the Game Awards, "4ÆM," is from Grimes' upcoming album, called "Miss Anthropocene," which is scheduled for release in February 2020. Grimes has so far released three singles off the album, including "So Heavy I Fell Through the Earth" and "Violence."

Source: Pitchfork

• In 2010, Goldman Sachs was charged $550 million by the federal government for misleading investors ahead of the collapse of the housing market. The settlement shows just the extent to which compliance violations can harm the public.

• Now, organizations are employing advanced technology to try to protect the public and prevent against future penalties.
• Anheuser-Busch InBev's purchase of rival SABMiller in 2015 created the need for a new tool to examine the company's compliance with US anti-corruption laws in the 23 countries where it operated. 
• Instead of going externally, global vice president for ethics and compliance Matt Galvin chose to build internally what eventually became BrewRIGHT. The platform allows Galvin to quickly review financial transactions across the globe for any potential illicit activity, helping to pinpoint where deeper investigations are needed.
• This article is part of Business Insider's ongoing series on Better Capitalism.
• Visit Business Insider's homepage for more stories.

In 2010, Goldman Sachs was charged $550 million by the federal government for misleading investors ahead of the collapse of the housing market. The bank was one of several that faced stiff penalties for similar actions.

The settlement shows the extent to which compliance violations can harm the public— nearly $8 trillion in value was lost from the stock market during the financial crisis — and the harsh fines that companies can face for breaches. Now, organizations are employing advanced technology to try to protect the public and prevent against future penalties.

When Anheuser-Busch InBev purchased rival SABMiller in 2015 for $100 billion, for example, the brewery giant knew it needed a new tool to review the company's adherence to US anti-corruption laws.

And to accomplish that, global vice president for ethics and compliance Matt Galvin turned to artificial intelligence and machine learning. Galvin worked with Tassilo Festetics, who as vice president of global solutions at AB InBev oversees its tech development efforts, and others to create BrewRIGHT, a platform that quickly reviews financial transactions across the globe for any potential illicit activity.

The program assigns a risk criteria based upon, among other things, the political connections of a specific vendor and whether that relationship could raise bribery or other ethical concerns, giving the compliance department much faster insight into whether certain transactions need to be investigated further.

"It was actually ultimately cost effective to do it this way," Galvin told Business Insider. "It solved the actual problem of the compliance integration, rather than this problem of a due-diligence exercise."

The application shows how AI and other advanced tech is permeating all aspects of business, even departments like legal that are typically less quick to adopt the latest and greatest technology.

Typically, compliance departments would review random samples of transactions across the countries in which it operates to review for any illegal payments. Galvin wanted to go further and take a more holistic view of all the information Ab InBev had on the 23 different countries that SABMiller operated in.

So he set out to centralize all the data from the firm and run the algorithm on that information to more quickly examine the transactions. When the tool was first deployed, it had a 99% failure rate. But as the organization better cataloged the data in a cloud-based repository, the accuracy improved.

"Data is always complicated at that level, because you have so many different sources, companies are not integrated," Festetics said.

Eventually, the team was able to expand the tool beyond just the SABMiller acquisition to monitor compliance across the entire company. Ultimately, it helps ensure that AB InBev adheres to laws intended to prevent corporations from using illicit methods — potentially at the expense of the general public — to further their own business interests.

And now, Galvin is working on creating a global consortium of sorts that would aim to pool resources to better combat compliance violations. 

Building the platform internally

For many companies pursuing AI-based initiatives, a key impediment to success can be resistance internally — like convincing middle-management to use the more advanced technology to improve operations.

AB InBev is no different. So it took a more targeted approach to start, beginning with projects that were unlikely to brush up against organizational challenges. BrewRIGHT was an ideal candidate because it addressed a major challenge facing the compliance department during a time in which much of the company was under pressure to integrate SABMiller's assets.

AB InBev chose to build it in-house, as opposed to working with outside partners. The decision gave the company control of the product and allows it to continually improve the platform without constantly having to go externally.

"It's very hard for you to maneuver if you always have a different partner that doesn't know your ecosystem," Festetics said. "It's very hard for you to continuously develop and to continuously build something useful for your company."

SEE ALSO: Accenture's head of artificial intelligence shares the 4-step plan every company should consider before investing in AI

Join the conversation about this story »

• Emergence Capital's first investment was in a pre-IPO Salesforce, and the firm has made smart bets since on companies like Zoom and Gusto. Now Jake Saper, partner at Emergence, is ready to make a new bet on the future of cloud software. 
• Saper thinks the next set of cloud software companies will be what he calls "coaching networks"— basically software that uses artificial intelligence to gather information on how you perform your daily tasks and provide suggestions for improvement and efficiency.
• The key is that AI will be used to augment, rather than automate, the tasks that the software is helping users do, he said.
• "It's a very different version or view of the way that AI will manifest in the enterprise than what is typically happening today," Saper told Business Insider. 
• Saper said that while he thinks some tasks will be automated by software, there's a huge opportunity to use AI to help retrain workers and open up more job possibilities. 
• Click here for more BI Prime stories.

The first deal Jake Saper was involved with at Emergence was its investment in Zoom, the video conferencing startup that rose to prominence this year when it went public in April. Zoom's IPO valued the company at $9.2 billion — roughly 9 times its last private valuation. 

But this wasn't the first success story for Emergence. In fact, the firm's first investment ever was a secondary round in Salesforce, ahead of its IPO. Since then, the firm has continually invested in a number of successful cloud-computing companies tackling the enterprise space, including Gusto, Box, and SucccessFactors. 

Now, Saper and Emergence Capital at large are staking their bets that the next big thing in cloud software will be something he calls "coaching networks"— software that gathers data from your daily tasks and offers back suggestions and guidance to help you work more efficiently. 

Saper said phase one of cloud was moving on premise software to the cloud, as we saw with Salesforce or Box. Phase two of cloud, which is happening now, is a trend towards industry-specific cloud software, which we've seen with the rise of apps like Emergence-backed Veeva, making software for the pharmaceutical industry.

Phase three, he said, will be these "coaching networks," built using AI — not to automate, as we see in most modern uses of AI, but to assist. 

"It's a very different version or view of the way that AI will manifest in the enterprise than what is typically happening today," Saper told Business Insider. 

He gives the example of a customer service representative at a call center. While on a call, software running in the background could collect data about the customer and use it to guide the representative as they move to assist. That data would also be used to help other employees using the same software for the same tasks, he explained.

This is exactly the sort of thing that one of his firm's portfolio companies, called Guru, is doing. "The idea is that it's capturing all of those pieces of knowledge and then it's surfacing them where the worker works...a really important element to this is that you still layer on top of the current customer interaction," Saper said.

The key is that artificial intelligence is used to augment each individual task, rather than for full-on automation. Saper thinks this kind of augmentation is the way in which AI will manifest itself in the software space going forward. 

"Low level tasks, what we call static tasks, where the data you need to effectively complete the task is relatively static and bounded. We think those tasks will be automated," Saper said. "The idea that we're more excited about is all of the, what we call dynamic tasks that exist in the enterprise, these would be basically human to human tasks."

AI will still change workers' lives

Of course, that's easy to say, but with AI comes the potential for job loss. 

According to a report by the Brookings Institution earlier this year, roughly 36 million American jobs have a "high exposure" risk from automation, meaning that 70% of their job functions could be performed by machines. Some of the most at risk jobs are focused on manual labor tasks such as cooks, waiters, and clerical office workers.

Saper thinks that some of those kinds of jobs will go away, but that AI can also be used to retrain and reposition workers for the jobs that will remain. 

"If I've got technology that I know can help augment those workers and get them much more quickly up the learning curve, then I can broaden the pool that I'm hiring from, which gives folks who may not have done the job before in the knowledge economy, the opportunity to do this knowledge economy job for the first time because they have a real time coach," Saper said. 

Got a tip? Contact this reporter via email at pzaveri@businessinsider.com or Signal at 925-364-4258. You can also contact Business Insider securely via SecureDrop.

Join the conversation about this story »

NOW WATCH: Taylor Swift is the world's highest-paid celebrity. Here's how she makes and spends her $360 million.

• Algae are thriving in waters in record numbers thanks to rising temperatures, and are creating dead zones where a lack of oxygen suffocates fish and other marine life below.
• But like plants, algae absorb carbon dioxide to grow. 
• Researchers are betting this ancient microorganism could combat global warming by drawing out greenhouse gas from the atmosphere.
• Hypergiant, an AI company, has developed a bioreactor that uses algae to capture carbon dioxide. Afterward, the algae is harvested to be turned into animal feed or as an ingredient in consumer products. 
• Watch the video above, part of the AI.Revolution series, to learn more.
• Visit Business Insider's homepage for more stories.

There are only a few ingredients needed for algae to take over: carbon dioxide, light, and water.

The ancient microorganism is thriving thanks to record heat waves and fertilizers washed away into nearby waters. And that's bad news for pretty much everything else. 

But what if a fourth ingredient — artificial intelligence — could transform the gooey sludge from a growing pest into a tool to fight climate change?

A team of researchers at the AI technology company Hypergiant sees algae as a weapon that can be harnessed for our benefit. 

They recently built an AI-powered machine, the EOS bioreactor, that takes advantage of algae's ability to capture carbon dioxide through photosynthesis. They say that by optimizing the growing environment for algae, their bioreactor can draw as much carbon dioxide out of the air as an acre of trees. 

It's one example of the carbon-capturing technology scientists say is necessary to avoid the worst effects of climate change.

So far, no one has been able to prove such a project can be done economically and at scale. But although the EOS bioreactor is only about the size of a refrigerator, Hypergiant claims it can do just that.

As the researchers explained, the machine harnesses the natural process of photosynthesis to filter emissions and capture carbon. Inside, glass tubes bubble with water and algae that are grown in artificial light. The AI monitors and regulates the algae's growth.

The EOS bioreactor is completely powered by AI, according to Hypergiant's research and development director Daniel Haab.

"Our goal is to make it so that no one needs to monitor or maintain this machine when it's out in the field," he said.

After the reactor has done its job capturing carbon dioxide, the algae can be harvested and extracted into a dry film.

The dried algae can then be mixed into animal feed or fertilizer, or can be used as an ingredient in a growing array of consumer products, such as moisturizer and nutritional supplements.

"You can do a lot of things with it and you can grow it literally anywhere," Haab said.

In recent years, steel factories and coal-burning power plants have installed algae-based technology to capture greenhouse gases before they are released into the atmosphere. Hypergiant's innovation adds a layer of intelligence to the chemical process.

The company did not comment on how many bioreactors would be required to make a measurable impact on the climate or whether it will be commercially viable.  

Still, the CEO says the company is on a mission.

"'Tomorrowing today,' our tagline, is about delivering the future that we were promised that we've read about, that we've heard politicians talk about," Hypergiant CEO Ben Lamm said. "Delivering on that future that isn't just aspirational but actually should be here."

The idea is to bring a build-it-yourself model to the masses. 

Haab hopes this DIY model will be done by releasing an open-source plan. He said Hypergiant is aiming to create a "connected grid of bioreactors that are all learning from each other, feeding each other information."

Although the researchers' goal is ambitious, one scientist said we're still a long way away from a long-term climate solution.

"In my opinion, this sort of approach is not going to sequester carbon at a rate anywhere near high enough to counteract climate change," said Kevin Flynn, a marine biologist at Swansea University in Wales.

Hypergiant's climate adviser Noam Bar-Zemer acknowledges there aren't "silver bullets" for the massive issue of climate change. 

"Problems this size get solved when innovators from around the world can take what they built, can bring those ideas together, and stitch those ideas into a tapestry of solutions that together, collectively, is big enough to meet the challenge," Bar-Zemer said.

So while a technological solution for carbon capture remains an elusive holy grail, these researchers hope to prove that artificial intelligence can help crack the code. Lamm is hopeful for the future, but said he still wishes people were more focused on the "benefits that society could bring, versus the 'we're all going to be dead in 50 years.'"

"Because who wants to wake up and do anything, if that's the case?" he said.

SEE ALSO: Most plastic is not getting recycled, and AI robots could be a solution

DON'T MISS: Farmers are using AI to spot pests and catch diseases — and many believe it's the future of agriculture

Join the conversation about this story »

• Descartes Labs uses artificial intelligence to detect wildfires, and they can correctly spot one in a record-breaking time of nine minutes.
• Devastating wildfires are occurring much more frequently than ever before. The earlier one can be detected, the better.
• Descartes Labs launched their wildfire detector on July 1, 2019, and helped the LA Times break news of the Kincade Fire.
• They leverage two satellites called GOES-16 and GOES-17 that hover above the western hemisphere and send thermal infrared images every five minutes.
• They are currently conducting a trial run with the New Mexico state government and hope to expand their services to all over the world.
• This article is part of our ongoing series on Better Capitalism.
• Visit Business Insider's homepage for more stories.

Nine minutes. That's how fast Descartes Labs can spot a growing wildfire with artificial intelligence. It's the record speed, and the current one to beat.

Normally, the process of detecting wildfires can take hours or longer. In the past, wildfires have been reported by civilians, commercial pilots, or fire agencies, who use crumbling towers built more than a century ago to look for flares. More recently, fire managers have started conducting reconnaissance flights to search for potential fires. But, since these are incredibly costly, this option is only employed when the fire risk is high.

Even when fires are spotted using one of these methods, it can take a long time to determine precisely where it is and how to get there. And with wildfires, there's no time to spare. The faster it spreads, the more dangerous it becomes. The effects can be catastrophic.

SEE ALSO: How $132 billion brewery giant AB InBev is using AI to fight corruption and spot business fraud around the globe

The rise and impact of wildfires

Examples include the Camp Fire, which burned more than 153,000 acres in Butte County, CA and killed 85 people in November 2018, and the October 2017 Tubbs Fire in Northern California, which resulted in 22 deaths. The 2014 Carlton Complex Fire ravaged 250,000 acres in Washington State. Fortunately, no one lost their lives, but hundreds of people were displaced.

These are just a few of the many recent wildfires destroying homes and habitats across the US. Until about a decade ago, these devastating fires only occurred in California, for example, once every 10 years — or less. Now, thanks to climate change, they happen once, twice, even three times every single year.

The simple fact is, the more quickly a wildfire is discovered, the better. Fire managers have more courses of action to choose from with smaller fires. The bigger it gets, the harder it is to fight, and the harder it is to evacuate people in time.

What Descartes Labs is doing to help stop wildfires

That's where Descartes Labs comes in, a startup launched in 2014 as a spin off of the government-run Los Alamos National Lab, Descartes Labs has gone through five funding rounds since May 2015, successfully raising a total $58.3 million, with the most recent series B round securing $20 million in October 2019.

On July 1, 2019, Descartes Labs officially launched its wildfire detector, which leverages satellites and AI to spot these dangerous, life-threatening fires. 

The idea originated during the summer of 2018. That year, New Mexico was experiencing extreme drought conditions due to an uncharacteristically low level of snowfall the previous winter. As a result, wildfires popped up left and right.

"You really got the sense that [the fires] were bearing down on you," Clyde Wheeler, an applied scientist at Descartes Labs, told Business Insider.

Wheeler and some of his coworkers would often go running together in the mountains near their Santa Fe headquarters. One day, they came to the peak of a mountain and saw a huge blaze off in the horizon.

"That, I think, really got us moving," Wheeler shared. "We were already sitting on a whole bunch of satellite imagery, and we'd already developed this AI technology, so we decided to see what we could do to help. That ended up being this real-time wildfire detector."

Leveraging existing satellites

Descartes Labs' wildfire detector takes the images from two geostationary satellites, GOES-16 and GOES-17 (also known as GOES East and GOES West), launched by the National Oceanic and Atmospheric Administration (NOAA) in 2016 and 2018, respectively. GOES-16 and GOES-17 hover above Earth, providing a stream of high-resolution and infrared imagery every five minutes.

"These satellites see into the thermal infrared"— meaning they can show the Earth's temperature — "and since the fires are really hot, hotter than the surrounding area and hotter than any point in the recent past, they appear as really bright, distinct spots. They're going to stand out," Wheeler explained. 

"Once we get an image," Wheeler said, "we start doing detection on it, start looking for fires in them. It takes nine minutes, from the point of the satellite capturing the image in space to us sending out an alert" if there is indeed a wildfire detected.

Each satellite remains above the exact same point on Earth at all times, meaning the pictures they take will always be of the same geographical areas. While GOES-16 and GOES-17 are positioned over different regions, their coverage overlaps. Together, they capture what's going on from New Zealand to the western coast of Africa, and from the Antarctic Circle to close to the Arctic Circle.

How AI comes into play

In order to ensure this tool correctly identifies a wildfire — and not something else — they had to create several different AI algorithms to filter out false alarms, like flares from the oil and gas industry, controlled agricultural burning by farmers, or giant fireworks shows, such as the one they saw in Fort Worth, TX this past Fourth of July. 

The algorithms take into account a number of different factors: where big oil and gas refineries might be, the location of steel and copper mills, where wildfires typically start, the overall terrain, and more. They must all be in agreement that the bright spots they're detecting are, in fact, a wildfire. 

One example of a possible false alarm was the 2018 Queens power plant explosion. The lights at LaGuardia Airport went dark, the sky turned a bright neon blue, and people wondered if the world was ending. While the satellites didn't know the answer about the fate of the planet, they did confidently declare that this hot spot was not a wildfire.

"It was actually a cool test [for the detector]," says Wheeler. "You could see it in the imagery, but our algorithms didn't detect a wildfire. "When the tool does spot a wildfire, Descartes Labs immediately broadcasts an alert to the LA Times who, in turn, reports it on their wildfire map. The detector can provide the exact latitude and longitude of the fire, which is crucial to taking action. Just a few short months ago, they did exactly this when they detected what's now known as the Kincade Fire in Sonoma County, CA, prompting the LA Times to release the first public notice about it.

The future of the Descartes Labs wildfire detector

So, what's next for Descartes Labs and their wildfire detector? In addition to alerting the LA Times, it's running a trial program with the New Mexico state government. Every time Descartes Labs detects a fire, the system sends a text alerting the state's fire managers, pinpointing the location and directions on how to get there.

The team wants to get this invaluable information into the hands of as many fire managers as possible, starting with the United States, then expanding to the entire area that GOES East and West cover. Eventually, they'd like to integrate images from other satellites, so they can help prevent incidents like the bushfires in Australia.

"It really is a project to do good," said Wheeler. "So we want to make sure that it's impact is maximized."

Wheeler rejects the view that AI has the potential to replace the human factor. The wildfire detector, he said, is here to, he said, "augment existing methods, to cast a better net that helps make sure fewer big fires slip through."

• Facial-recognition algorithms are more likely to misidentify people of color than white people, according to a federal study published on Thursday.
• The study found that black people and Asian people were up to 100 times as likely to produce a false positive than white men, and women were more likely to be misidentified than men across the board.
• Law enforcement across the US is embracing facial-recognition technology as a tool for identifying suspects, a trend that the study calls into question.
• Some prominent facial-recognition software, like Amazon's Rekognition, was not made available for the study.
• Visit Business Insider's homepage for more stories.

A sweeping federal study of facial-recognition technology found that the systems were worse at identifying women and people of color than men and white people, the National Institute of Standards and Technology announced on Thursday.

Researchers found that facial-recognition software produced higher rates of false positives for black people and Asian people than whites. The software had a higher rate of false positives for those groups by a factor of 10 to 100 times, depending on which algorithms were used.

Women were also misidentified more frequently than men across the board, the study found. The Native American demographic had the highest rate of false positives.

"While it is usually incorrect to make statements across algorithms, we found empirical evidence for the existence of demographic differentials in the majority of the face recognition algorithms we studied," the NIST researcher Patrick Grother, the report's primary author, said in a statement. "This data will be valuable to policymakers, developers and end users in thinking about the limitations and appropriate use of these algorithms."

Law-enforcement agencies across the US have begun to embrace facial-recognition technology as a tool for tracking people's movement and identifying suspects in the past year. Civil-liberties advocates have pushed back against that trend, and some cities have banned the use of facial recognition by law enforcement.

The NIST study confirms existing research that has shown racial and gender bias of facial-recognition technology. Researchers at MIT found that Amazon's facial-recognition software, Rekognition, misidentified people of color at a higher rate than white people (Amazon criticized the study, arguing that researchers were using the software incorrectly).

Amazon was not one of the 99 facial recognition vendors tested by NIST, however — the company did not make its software available for the study, according to The Washington Post. Vendors who participated in the study included Intel, Microsoft, Panasonic, SenseTime, and Vigilant Solutions.

An Amazon spokesperson declined to comment. In the past, Amazon has said its software is a cloud-based service and would need to be altered in order to undergo NIST's test.

The NIST study also found that facial-recognition software made by Asian companies was less likely to misidentify Asian faces.

"These results are an encouraging sign that more diverse training data may produce more equitable outcomes, should it be possible for developers to use such data," Grother said in a statement.

SEE ALSO: How police are using technology like drones and facial recognition to track people across the US

Join the conversation about this story »

NOW WATCH: Watch Google reveal the new Nest Mini, which is an updated Home Mini

• Facebook and Twitter dismantled thousands of accounts and pages that were part of a global network seeking to push pro-Trump messaging. 
• NBC News reported that the Facebook accounts had been identified as having ties to the Epoch Times, a pro-Trump conspiracy website. 
• The group was previously second only to the Trump campaign itself in pro-Trump ad spending with $1.5 million before it was banned in August. 
• Visit Business Insider's homepage for more stories.

Facebook reportedly removed more than 600 accounts that were tied to a pro-President Donald Trump conspiracy website apparently used to push conspiracies and fake news. 

NBC News reported that the Epoch Times used artificial intelligence to push a variety of political stories through a network run by Vietnamese users pretending to be Americans on the site in addition to 89 pages, 156 groups, and 72 Instagram accounts that amassed over 55 million followers.

The network also appeared on Twitter, where a spokesperson told the Wall Street Journal the site suspended approximately 700 accounts and was undergoing investigations into their origins. 

Facebook's head of security policy, Nathaniel Gleicher, told NBC News that the company observed what they believed to be "a US-based media company leveraging foreign actors posing as Americans to push political content."

"We've seen it a lot with state actors in the past," Gleicher told the outlet. 

The accounts pushed "anti-impeachment" and "pro-Trump" messages with AI-generated faces atop their profiles, which Gleicher said can be detected by the site's automatic monitoring for fake accounts. 

This is the latest run-in with Facebook the pro-Trump conspiracy website since August, when it was banned from buying ads after it attempted to bypass its review system. Before the ban, the Epoch Times was one of the largest buyers of pro-Trump ads on Facebook, spending more than $1.5 million on pro-Trump Facebook advertising throughout 2019, second only to the Trump campaign itself. 

Stephen Gregory, the Epoch Times' US editions publisher, said in a statement after the site announced removing the accounts that the media group has no connection to the fake network, and blamed some former employees for the "unjustified ban." 

The Epoch Times' official Facebook pages are still active and "verified" on Facebook, and its main account has nearly 6 million likes. 

The Epoch Times is a non-profit publication that's tied to a Chinese religious group named Falun Gong, which seeks to undermine the Chinese government and sees Trump as a useful tool. 

The revelation of the network comes as Facebook is under mounting scrutiny for the abundance of fake news and ads that thrive on the site, which is an especially concerning factor ahead of the 2020 US presidential election. 

SEE ALSO: The 10 most-viewed fake-news stories on Facebook in 2019 were just revealed in a new report

Join the conversation about this story »

NOW WATCH: Extremists turned a frog meme into a hate symbol, but Hong Kong protesters revived it as an emblem of hope

• LinkedIn released its annual report of the 15 top emerging jobs in 2020.
• These jobs have seen high annual growth over the last five years.
• Many jobs — including full stack engineer and back end developer — are in engineering.
• Visit Business Insider's homepage for more stories.

Artificial intelligence experts are in luck in 2020.

LinkedIn released its list of the top emerging jobs for 2020. These jobs have grown substantially in the last five years, and LinkedIn predicts they will continue to increase demand in the new year.

Demand for artificial intelligence specialists grew 74% over the last five years. The job requires fluency in deep learning and machine learning. Cities hiring the most for artificial intelligence specialists include Boston and San Francisco. 

Engineering roles overwhelmingly made up the top 15 jobs, but roles like customer success specialists and chief revenue officers will also be in demand next year.

Here are the 15 top emerging jobs of 2020, according to LinkedIn.

SEE ALSO: These were the 50 best US companies for women to work for in 2019

15. Product owners have seen 25% annual growth in demand in the past five years.

Skills unique to the job: Agile methodologies, scrum, product management, software development, JIRA

Top industries hiring this talent: Information technology and services, financial services, computer software, insurance, hospital and healthcare

Top hiring cities: New York, Boston, San Francisco Bay Area, Chicago, Atlanta 

14. JavaScript developers have seen 25% annual growth in demand in the past five years.

Skills unique to the job: React.js, Node.js, AngularJS, JavaScript, cascading style sheets 

Top industries hiring this talent: Computer software, information technology and services, internet, financial services, marketing and advertising

Top hiring cities: New York, San Francisco Bay Area, Los Angeles, Boston, Washington, D.C.

13. Cloud engineers have seen 27% annual growth in demand in the past five years.

Skills unique to the job: Amazon Web Services, cloud computing, Docker products, Ansible, Jenkins

Top industries hiring this talent: Information technology and services, computer software, financial services, internet, telecommunications

Top hiring cities: San Francisco Bay Area, Washington, D.C., New York, Dallas-Fort Worth, Chicago

12. Chief revenue officers have seen 28% annual growth in demand in the past five years.

Skills unique to the job: Strategic partnerships, startups, software as a service (SaaS), go-to-market strategy, executive management

Top industries hiring this talent: Computer software, information technology and services, marketing and advertising, internet, financial services

Where the jobs are: New York, San Francisco Bay Area, Atlanta, Boston, Chicago

11. Back end developers have seen 30% annual growth in demand in the past five years.

Skills unique to the job: Node.js, JavaScript, Amazon Web Services, Git, MongoDB

Top industries hiring this talent: Computer software, internet, information technology and services, marketing and advertising, financial services

Top hiring cites: San Francisco Bay Area, New York, Los Angeles, Boston, Seattle

10. Cybersecurity specialists have seen 30% annual growth in demand in the past five years.

Skills unique to the job: Cybersecurity, information security, network security, vulnerability assessment

Top industries hiring this talent: Information technology  and services, defense and space, computer network and security, management consulting, financial services

Top hiring cities: Washington, D.C., New York, San Francisco Bay Area, Chicago, Denver

9. Behavioral health technicians have seen 32% annual growth in demand in the past five years.

Skills unique to the job: Applied behavior analysis, autism spectrum disorders, behavioral health, mental health

Top industries hiring this talent: Mental healthcare, hospital and healthcare, individual and family services, education management, health, wellness and fitness

Top hiring cities: Miami-Fort Lauderdale, Orlando, San Francisco Bay Area, Phoenix, Seattle

8. Data engineers have seen 33% annual growth in demand in the past five years.

Skills unique to the job: Apache Spark, Hadoop, Python, Extract/Transform/Load (ETL), Amazon Web Services

Top industries hiring this talent: Information technology and services, internet, computer software, financial services, hospital and healthcare

Top hiring cities: San Francisco Bay Area, New York, Seattle, Boston, Chicago 

7. Sales development representatives have seen 34% annual growth in demand in the past five years.

Skills unique to the job: Salesforce, cold calling, software as a service (SaaS), lead generation, sales

Top industries hiring this talent: Computer software, internet, information technology and services, marketing and advertising, computer and network security

Top hiring cities: San Francisco Bay Area, New York, Boston, Chicago, Austin

6. Customer success specialists have seen 34% annual growth in demand in the past five years.

Skills unique to the job: Software as a Service (SaaS), Salesforce, customer relationship management, account management, customer retention

Top industries hiring this talent: Computer software, internet, information technology and services, marketing and advertising, financial services

Top hiring cities: San Francisco Bay Area, New York, Boston, Chicago, Washington, D.C.

5. Site reliability engineers have seen 34% annual growth in demand in the past five years.

Skills unique to the job: Amazon Web Services, Ansible, Kubernetes, Docker products, Terraform

Top industries hiring this talent: Internet, computer software, information technology and services, financial services, consumer electronics

Top hiring cities: San Francisco Bay Area, New York, Seattle, Boston, Washington, D.C.

4. Full stack engineers have seen 35% annual growth in demand in the past five years.

Skills unique to the job: React.js, Node.js, JavaScript, AngularJS, cascading style sheets 

Top industries hiring this talent: Computer software, information technology and services, internet, financial services, higher education

Top hiring cities: San Francisco Bay Area, New York, Los Angeles, Boston, Washington, D.C.

3. Data scientists have seen 37% annual growth in demand in the past five years.

Skills unique to the job: Machine learning, data science, Python, R, Apache Spark

Top industries hiring this talent: Information technology and services, computer software, internet, financial services, higher education

Top hiring cities: San Francisco Bay Area, New York, Washington, D.C., Seattle, Boston

2. Robotics engineers have seen 40% annual growth in demand in the past five years.

Skills unique to the job: Robotic process automation, UiPath, Blue Prism, Automation Anywhere, Robotics

Top industries hiring this talent: Information technology and services, industrial automation, computer software, financial services, automotive

Top cities hiring: San Francisco Bay Area, Atlanta, New York, Washington, D.C., Boston

1. Artificial intelligence specialists have seen 74% annual growth in demand in the past five years.

Skills unique to the job: Machine learning, deep learning, TensorFlow, Python, natural language processing

Top industries hiring this talent: Computer software, internet, information technology and services, higher education, consumer electronics

Top cities hiring: San Francisco Bay Area, New York, Boston, Seattle, Los Angeles

• The cloud market has grown rapidly in the last two decades, and Rory O'Driscoll, partner at Scale Venture Partners, thinks the market is about to experience a slowdown. 
• As cloud increasingly becomes the norm, O'Driscoll thinks the next phase of cloud will have cloud companies competing with each other, rather than with legacy players who have on premise software, which has been the business model up until now.
• O'Driscoll has been investing in enterprise software for the past 25 years and actively invested in the software as a service trend from the beginning as enterprise software moved to the cloud.
• He lays out three strategies for startups to compete in this new environment. 
• Click here for more BI Prime stories.

The cloud market has been growing for the last two decades. It's led large companies like Amazon, Microsoft, and Salesforce to new highs and opened a huge opportunity for new players to disrupt the existing software market.

However, that growth is expected to slow in the next three to five years, as cloud increasingly becomes the norm, argues Rory O'Driscoll, partner at Scale Venture Partners. O'Driscoll thinks the next phase of the market will see cloud companies competing with each other, versus trying to take market share from on-premise software.

O'Driscoll has been investing in enterprise software for the past 25 years and actively invested in the software as a service trend from the beginning. He was an investor in cloud companies like Box, Docusign, and Bill.com ahead of their respective IPOs, and holds several board seats at such firms.

In a recent conversation, O'Driscoll told Business Insider that from his very close perch, a major tipping point in the industry is approaching: By 2021, he says, the majority of all software revenue across the industry will be derived from the cloud, rather than the more traditional approach.

At that point, he said, cloud companies will not be able to rely on the old model of taking market share from legacy vendors — they will have to compete directly with other cloud companies. 

"There will still be lots of opportunities to build, to take market share from on prem, but that will be decreasing and more and more it will be about competing with other cloud companies. That's going to be qualitatively harder to do," O'Driscoll said. 

In a blog post on the topic, O'Driscoll was a little more colorful: "Goodbye Gold Rush, hello Hunger Games."

How to succeed

The way for startups to compete successfully going forward is to build automation tech that simplifies the process of working across all of the very many cloud services required for getting a job done.

"The sweet spot will be going beyond the cloud, which is building on top of the cloud and focusing more on software that does the work for us and automates the work, and that leverages AI to not only move business processes from on premise to the cloud but to actually remove those business processes and do the work for us," O'Driscoll told Business Insider when we spoke in November. 

He thinks large cloud companies, like Salesforce, will start to feel the pressure when they can't grow anymore in the core markets they are in, and have to turn to new markets to keep growing. Companies in that position will start to bundle products together, in the same way Microsoft does when selling its software, he added. 

That creates an even greater challenge for startups looking to enter the space. Instead of competing with legacy players, startups will have to contend with cloud behemoths who are "in an adjacent market that wants to add your product and take your market,"O'Driscoll wrote in his blog post. 

This dynamic is already playing out with the competition between Slack and Microsoft Teams. Microsoft bundles its Teams workplace chat app for free in with its other Office 365 productivity software, and is therefore making it harder for Slack to gain market share and keep growing. 

Three possible roadmaps

O'Driscoll said he has three strategies for new startup founders to follow in what he calls this new, "harsher cloud" environment that will develop in the next five years. 

• The first is to enter an existing market and win against legacy players. One example of a success story with this strategy is video conferencing company Zoom, O'Driscoll said. Zoom took on legacy players like Cisco (WebEx), Microsoft, and LogMeIn. 
• The second is tacking an area where cloud software has not really caught on. Some of those areas are "many industry verticals, most SMB's, and the parts of the cloud that are being carved out of the wider telco industry," O'Driscoll writes. 
• The third is focus on building software that uses new technologies like artificial intelligence for automating tasks. This software can be built on top of cloud apps and is meant to let users do more with the software, O'Driscoll said. 

This last strategy, O'Driscoll said, is already happening in the workplace, and companies that can build in that space will be the best-positioned to compete in a crowded market.

Got a tip? Contact this reporter via email at pzaveri@businessinsider.com or Signal at 925-364-4258. You can also contact Business Insider securely via SecureDrop.

Join the conversation about this story »

NOW WATCH: 5 things about the NFL that football fans may not know