• Facebook's chief technology officer, Mike Schroepfer, has been in charge of trying to rid the social platform of false and dangerous content.
• In a recent interview with The New York Times, Schroepfer teared up while discussing the pressure he's under to fix the problem.
• Facebook has upped its efforts to censor offensive and violent content since the shootings in Christchurch, New Zealand, were livestreamed on the platform.
• Visit Business Insider's homepage for more stories.

The pressure Facebook has faced trying to eliminate violent and offensive content from the platform is enough to make a grown person cry — literally, if you ask Facebook executive Mike Schroepfer.

Schroepfer, Facebook's chief technology officer, teared up several times during a series of interviews with The New York Times about the platform's recent policing efforts. Criticism of the platform has ramped up since the terrorist attacks in March on Christchurch, New Zealand, during which the shooter livestreamed his attack on Facebook.

Schroepfer "choked up" when talking about "the scale of the issues that Facebook was confronting and his responsibilities in changing them," The Times reported.

"It won't be fixed tomorrow," Schroepfer said about Facebook's efforts. "But I do not want to have this conversation again six months from now. We can do a much, much better job of catching this."

Read more:Facebook is dialing up punishments for users who abuse live video after the Christchurch massacre

The CTO is known for "often" letting his feelings shows, "many" people told The Times. A former Facebook employee, Jocelyn Goldfein, a venture capitalist, said she'd seen Schroepfer cry in the office when she worked for the social platform.

Schroepfer has been tasked with building artificial-intelligence tools for Facebook that will better work to detect harmful content, and can prevent something like the Christchurch shooting from being broadcasted on Facebook again.

To figure out how Facebook's technology can best identify the next terrorist-related video, Schroepfer had to watch the gruesome footage of the shooting "several times," according to The Times.

"I wish I could unsee it," Schroepfer said.

Facebook has taken some steps to avoid an incident like the New Zealand shooting livestream from repeating itself. The platform has implemented a "one strike" policy that blocks users immediately from livestreaming if they violate Facebook's "most serious" rules.

The company has invested $7.5 million into research on better techniques for detecting videos that have been manipulated, which is how millions of repostings of the Christchurch shooting got past Facebook's automated system and spread online.

Schroepfer told The Times that his task of removing harmful posts is a complex one without an "endgame."

He said the number of posts is "never going to go to zero."

SEE ALSO: Google is scanning your Gmail inbox to keep a detailed list of your purchases, and there's no easy way to erase it

Join the conversation about this story »

NOW WATCH: This London handbag company has recycled 175 tons of fire hoses into fashion accessories

• London's Heathrow Airport is one of the busiest airports in the world, and only getting busier. But it uses only two runways to handle the 1,300 flights that take off and land every day. 
• Since Heathrow can't get any bigger, it had to get smarter, which is why it's using a number of Microsoft products and cloud services to apply artificial intelligence to the problem. 
• Heathrow's chief information officer says that its AI systems are improving passenger experience, cutting down on flight delays, and generally crunching an absurd amount of data to make everything more efficient. 
• It's also using Microsoft's PowerApps platform to help employees build their own apps that draw on the airport's data to help them do their jobs better and faster — while cutting down on paperwork. 
• Visit Business Insider's homepage for more stories.

Happy passengers, travelling with their bags, on time.

It sounds simple, but it's harder to deliver when you're the seventh-busiest airport in the world, handling record numbers of flights and far more passengers than Heathrow was ever designed for. That's why Heathrow Airport turned to the cloud and AI — mainly powered by Microsoft technology.

At Heathrow, 1,300 flights take off and land every day, on 84 airlines, going to 203 destinations in 84 countries. That's 80 million passengers a year, using just two runways, and a plane taking off or landing every 45 seconds. Plans for a third runway finally got government approval in 2018; even if the expansion goes smoothly it won't be in use until 2026, when it will push passenger numbers to around 130 million a year.

Even without that room to grow, the numbers of flights and passengers from Heathrow are creeping up; from 207,000 passengers a day to 220,000. 1300 flights a day almost reach the 480,000 takeoffs and landings the airport is allowed every year.

Since Heathrow can't get bigger, the airport team turned to Microsoft for help getting smarter. The airport's management boasts that by using AI tools that integrate with existing operational tools makes Heathrow the best large airport in Europe for getting flights out on time, despite being stuck between airport congestion in Europe and the fluctuations in the jet stream. 

Hundreds of different data sources feed into a model of everything that affects the airport: real-time data of flight movements, passenger transfers, security and immigration queues, boarding card scans, baggage handling, car parks, world events and even currency rates with weather predictions.  It uses SQL Azure database storage, Azure Stream Analytics — a real-time event processing service that can detect anomalies, predict trends and trigger workflows —Azure Data Lake Analytics to handle analytics queries, and the Power BI business intelligence service for dashboards that tell everyone working at the airport what's going to happen long before planes land.

"The scale of complexity we work at and the fact we're capacity constrained is actually really exciting because it forces us to be a bit creative," Heathrow CIO Stuart Birrell told Business Insider. "We have to look at how we operate to squeeze efficiency and effectiveness and make sure passengers get a great experience. There's a huge amount of disruption across the global aviation network and we're at the heart of that."

There are plenty of companies out there betting big on AI, which is expected to change just about every industry, everywhere: Amazon Web Services, the leading player in the cloud computing space, has made significant investment in that area, too. Birrell says that the airport did its research before settling on Microsoft, but that its existing relationship with the tech titan certainly played a role.

"The airport has been a Microsoft customer for years, and decided to go with Azure and Power BI only after a thorough commercial and technical review of the market and competitor products," Birrell says. 

Earlier isn't always better

It's not just delays; planes landing early can be just as much of a problem. Day to day changes in the jet stream mean flights from the US can reach Heathrow half an hour late or an hour early; flight times from the far- east change too, Birrell says. 

"We could have fifteen or twenty flights arriving 45 minutes early," says Birrell; each one of those disrupts another two or three other flights. When long haul flights arrive early, short haul flights from Europe and elsewhere in the UK lose their landing slots and have to be delayed.

As well as space on the runway, flights need a gate to land at, an airbridge to get people off the plane, and airport staff to deal with them. Twenty early flights mean 6,000 people arriving 45 minutes before immigration, customs, baggage handlers and the staff in the coffee shop are ready for them.

"The planes need handling too; they need engineering, they need fuel, they need cleaning and food restocking. It's a very complex world," Birrell explains.

Heathrow's AI models aren't just predicting when planes will land; it covers everything from when passengers will get to the airport and go through security to what happens behind the scenes, and even after their flight has left.

Birrell says that the system takes a lot of things into account:

"We have simple things like passenger food carts on flights by airline. We have the scheduling and what's likely to happen with air traffic control. We work on weather patterns; we know what the jet stream is going to do 12 to 24 hours ahead of time. Our operations center actively monitors social media for incidents, we monitor roads and railways systems around the south of England, we monitor what's going on across Europe. In the last 48 hours we stopped flying over Pakistan; that impacts flight schedules across the middle east. We feed all that in with the rostering and resource planning constraints we have. It's building into a very sophisticated model but we're trying to break it down into something manageable and understandable."

The predictions made by the model go to 20,000 staff across 400 different companies at the airport who need operational and scheduling data, Birrell says. Knowing how many people will be coming through immigration and when means the UK Border Force can staff enough desks. It predicts when aircraft will need re-fuelling or towing away from the gate – and it does it far enough in advance to staff to have a plan rather than just reacting to what's already happened.

"If a plane is going to be arriving late in two hours' time, if we know about it happening several hours in advance, if it's not a surprise, we can make it a much smoother experience. We can get buses out to certain aircraft to expedite transfers or rebook people automatically onto other flights," Birrell says. 

Predictions, not paper

Departing flights aren't all the same either.

"If a flight is leaving at 1 p.m., not all passengers are going to get to the airport two hours early; some will get there three hours early, some will only get there half an hour before and some will be coming from other flights. People behave differently depending on what destination they're flying to."

The different patterns of passengers getting to the airport can cause queues for check-in and security; predicting passenger flow helps staff in the airport stores and departure lounges too. "We have 4,000 security officers and we need to plan the roster to have the right number of security lanes open to avoid queues."

Planning security shifts used to be done using spreadsheets; now the same team is using Azure Machine Learning  — a managed cloud service that combines popular data science tools with automated systems that suggest the best machine learning algorithm to use — to predict how many passengers will be coming through the security line at every terminal in 15-minute intervals.

Managers see the forecast in Power BI dashboards so they can plan where to put staff in real time, rather than sticking to a fixed schedule, meaning that passengers are less likely to get stuck in a long security queue than they were two years ago.

"Security is the biggest stress point for passengers; if we can provide a great service then they're less stressed as they come through the airport," Birrell says. 

Coding without coding

Working a security lane shift can be stressful too, and it usually involves a lot of paperwork: everything from shift schedules to reports to assisting travellers. Samit Saini spent 13 years working security at Heathrow; looking at X-ray machines and asking passengers if they had any liquids in their bags. He was frustrated by how inefficient it was to pull out the tattered sheet of paper with translations of standard phrases into different languages every time a passenger didn't understand him.

So he taught himself to use PowerApps, Microsoft's 'low code' service for making your own smartphone apps without needing to be a developer, and made an app for that. Now his job is making apps using the giant data model of the airport; one app he's about to deploy will get rid of 30,000 pieces of paperwork security officers and managers have had to fill out every year, by moving to electronic signatures and auditing.

Birrell is as excited about that as the security staff.

"If we can create an environment where security officers have all the data and they're not worried about paperwork because it's all happening in the background, they can provide a much better experience for the passenger so the passenger is less stressed and everyone is a winner," Birrell says. 

Other areas of airport operations will also go paperless, leaving staff to concentrate on doing their job rather than filling out a form that says they did it. The amount of airport data in the Microsoft Azure cloud that can be used for analytics and predictions is about to triple, and teams or individual employees can use it make their own Power BI dashboards and PowerApps when they need them.

It's like the next generation of Microsoft Office, Birrell enthuses.

"If you understand Excel, you can use these tools and get really exciting output and results. You don't have to ask IT to run a report for you or do a piece of analysis for you; you can do it yourself," he says. 

Having an AI model that tells you what's going to happen can beat congestion and get flights out on time by turning planes around quickly – but only if the information gets to the right people around the airport.

Everything about your trip through the airport gets better because airport staff know what's going to happen and they're prepared for it. That keeps Heathrow consistently in the top ten airports for passenger satisfaction worldwide, and why that score keeps going up even though it's operating at capacity, Birrell says.

And when the third runway does arrive, the team at Heathrow plans to use what they've learned from the Azure AI model about how passengers move around the airport and through security to design new flows around the airport, without the disruption expansion usually brings.

Join the conversation about this story »

NOW WATCH: WATCH: The legendary economist who predicted the housing crisis says the US will win the trade war

• Many Chinese tech firms are equipping the country's farms with artificial intelligence, facial recognition, and blockchain technologies to monitor their pigs' health.
• But none of this has managed to curb the biggest epidemic facing Chinese farms today — the outbreak of the African swine flu, a gruesome and deadly disease that's poised to leave 200 million pigs dead.
• It shows the limit to China's technological ambitions, which many observers consider dystopian and Orwellian.
• China is the world's largest consumer and producer of pork, and the swine flu is expected to affect the world's supply, demand, and price of pork.
• Visit Business Insider's homepage for more stories.

China is rolling out artificial intelligence, facial recognition and blockchain technology across its rural farms to protect its pigs — but it still isn't enough to curb a deadly swine flu that's spread across the country and is killing a third of its hogs.

Tech firms, from small to large, have been developing various technologies to monitor the health and whereabouts of Chinese pigs. China is the world's largest pork producer and consumer.

Chinese tech giants like e-commerce company Alibaba, its rival JD.com, and gaming company NetEase have already been developing advanced technologies to monitor the health and whereabouts of the country's pigs, The New York Times and The Daily Beast reported.

Alibaba has already developed voice recognition software to monitor pigs' coughs as well as cameras to monitor hogs' faces, the Times reported in February.

AI systems on some farms determine how fast each pig is growing, and monitor whether it is healthy. Farms can then use that data to control the animals' feed and environmental factors, the Daily Beast reported.

The sound sensors pick up on pigs' coughs and flag potential illnesses to farm operators, it added. Some farmers are also using blockchain technology to track pigs from their birth all the way to their slaughter and selling in the market.

Here's how Jackson He, the CEO of Yingzi Technology, small facial and voice recognition company, sketched out his vision of a "future pig farm" to the Times: "If they are not happy, and not eating well, in some cases you can predict whether the pig is sick."

But the advanced technologies have still not managed to curb the biggest epidemic facing Chinese farms today — the outbreak of the African swine flu (ASF), a gruesome and incurable disease afflicting the country's pigs.

Pigs that contract ASF can experience high fevers, diarrhea, depression, and bluish skin, and pregnant sows can lose their babies, according to the World Organisation for Animal Health. Animals with acute ASF typically die within days.

Some 1.02 million Chinese pigs have already been culled this year because of the disease, according to the agricultural ministry. The country is expected to kill about a third of all its pigs this year alone, Dutch bank Rabobank estimated last month.

The country is expected to produce 150 million to 200 million fewer pigs, amounting to a 25% to 35% drop in production, Rabobank said. This fall in pork production is likely to push up the global demand for pork, which could result in supply shortfalls and price rises.

Read more: China is killing a third of its pigs because of a gruesome and incurable fever, which could drive up the price of pork around the world

Beijing is desperately trying to stem the outbreak — which has spread to almost the entire country, and to Vietnam and Cambodia — by setting up buffer zones around infected areas, making slaughterhouse inspections more stringent, and threatening to punish farmers who don't report outbreaks.

But many farmers and producers are still failing to detect infections and aren't taking the correct safety precautions, with many adding that local officials were taking too long to respond to potential outbreaks, the Times and CNN reported.

Some farmers have also been disposing of their pigs in an unhygienic way, in some cases burying dozens of them alive, or piling them in rivers and ditches.

Limits of China's 'Big Brother' technology

The ASF epidemic is an example of how, despite the Chinese government's grand claims about its technological advances, there are still severe limits to what it can achieve.

The country has over the past few years poured billions of dollars into AI, facial recognition, and blockchain research, with experts saying that the country's output of AI research could soon overtake the US.

Many news outlets have likened China's technology, which the government claims is all-seeing and all-knowing, to "Big Brother," the all-seeing government in George Orwell's novel 1984.

Read more:In China you get a special warning before you call people who owe money, telling you to get them to pay up

But much of the technology remains fairly limited and unreliable at this point.

The country has dozens of datasets and systems containing citizens' — and perhaps even animals' — personal information, but many of the officials tasked with updating the systems are often too lax or overworked to maintain them, Human Rights Watch senior China researcher Maya Wang wrote in The Wall Street Journal.

Data sharing across local governments also remains a problem in China, with some ministries and municipalities distrusting each other so much that they won't share information, New York Times China tech reporter Paul Mozur noted earlier this year.

President Donald Trump's administration is reportedly considering a ban on exports of US components to Hikvision, a giant Chinese surveillance company.

SEE ALSO: China's 'Big Brother' surveillance technology isn't nearly as all-seeing as the government wants you to think

Join the conversation about this story »

NOW WATCH: We tried a fermentation-tracking device and highly recommend it to find out which foods are making you bloated

• TouringPlans.com has been using artificial intelligence to help users plan the perfect vacation at Walt Disney World, Disneyland, and other major tourist destinations, for subscriptions starting at $16 per year.
• Punch in the dates of your trip, and it tells you projected crowd levels at the Disney parks. Enter your proposed itinerary of rides, restaurants, and shows, and it uses AI to come up with an optimized itinerary that minimizes time spent in line. 
• Founder Len Testa tells Business Insider that TouringPlans.com has some 140,000 paying customers.
• The site was built based on complex algorithms developed while Testa was in a graduate school, but has evolved into so much more since. 
• Visit BusinessInsider.com for more stories.

We're not going to Walt Disney World in Orlando, Florida, for another six months, but I'm already getting neurotic about planning the trip.

How neurotic? I have an alarm set for 3 a.m. on Sunday morning, which is the earliest that we can start making restaurant reservations for our trip. There's even a spreadsheet with all the restaurants we want to visit, as meticulously researched from videos and handy guides from places like Disney Food Blog and WDW Prep School.

So I was more than receptive when someone recommended that I check out TouringPlans.com— a site that uses complex algorithms to help you plan the perfect vacation at Disney World, Disneyland, or a handful of other theme parks like Universal Studios Orlando. It's a premium service, with subscriptions priced starting at $16 per year. 

Len Testa, the founder and president of TouringPlans.com, tells Business Insider that the service currently has 140,000 paying users, but that it's had millions of users total since its inception in 2012. It has seven full-time employees on the staff, and 12 part-time employees, says Testa, as well as a business in publishing "Unofficial Guides" to Disney World, Disneyland, and other popular tourist destinations.  

The service is important, says Testa, because with a little forward planning, you can make the most out of your very limited vacation time at Disney World. Even if you only save an hour a day, that's an hour that can be spent going on your favorite ride again, or getting a better seat for the fireworks. 

"It's so important to have a plan," says Testa. "An hour is a pretty good investment." 

What TouringPlans.com does

After playing with it for a few days, I can tell you this: TouringPlans.com may not be much to look at, but much like the Millennium Falcon — the star of this year's new Star Wars: Galaxy's Edge lands at both major Disney parks— it's got it where it counts. 

Read more: 5 things to know before riding Millennium Falcon: Smuggler's Run in Disneyland's Star Wars: Galaxy's Edge

TouringPlans.com helps you decide which Disney World park to visit by analyzing historical attendance data to predict crowd levels on any given day. Then, for each day of your visit, just plug in all the rides, shows, and restaurants you want to hit up, and it spits out an optimized itinerary that promises to minimize wait times while still getting you everywhere you want to go.

It even makes recommendations on how, when, and where to deploy FastPass+, which lets you make reservations to go on the resort's most popular rides ahead of time.

But wait, there's more: It'll alert you if hard-to-get reservations at in-demand eateries like Be Our Guest at the Magic Kingdom open up. A personal favorite feature is that it has a directory of the view from every single Disney World resort hotel, and can automatically send a fax to the front desk requesting a specific room. The site even recommends rooms to request, based on criteria like distance from the food court or time to the bus stop. 

It started in line

Fittingly, the idea that would become TouringPlans.com struck Testa while he was in line at Disney World during a family trip in 1997 — specifically, during a two-hour line for the now-defunct Great Movie Ride at Disney-MGM Studios (now known as Disney Hollywood Studios).

"In the middle of the two hours, I thought, 'there has to be a better way to do this,'" says Testa. 

At the time, Testa had just finished his undergraduate degree at North Carolina Agricultural and Technical State University, and was getting ready to re-enroll at the school for a grad program in computer science. As soon as he could, he took his big idea — using algorithms to improve the theme park experience — to his advisors. 

They had two questions, he said. First, "besides you, does anyone care?" And secondly, they questioned if the problem was hard enough to be worth Testa's time to solve. It turned out that, yes, it was: Planning the perfect Disney vacation is actually very similar to the travelling salesman problem, a famously complex computer science problem that requires the calculation of the optimal route between a set of cities. 

As his graduate thesis, Testa wrote the core statistical models that would grow into TouringPlans.com. Today's version, he says, is three or four generations removed from that original project. The fact that it only needed that many revisions over the 20 or so years since inception is proof that the foundations were solid, he says. 

Nowadays, companies like Google, Microsoft, and Amazon use similar approaches as Testa's original algorithms, as Silicon Valley increasingly turns to artificial intelligence to solve complex problems. At the time that TouringPlans.com was conceived, though, AI was still a long ways off from its current moment in the spotlight.

"I think we were a little ahead of the curve there," says Testa.

And, as a side-note, in 2015 Testa would later team up with Dr. Bradley Eilerman — a fan of TouringPlans.com  — to launch GlucosePATH, a service that uses modified versions of these very same algorithms to help diabetes patients manage their medication. That's currently in a pilot program with Kentucky Medicaid, says Testa.

A data-driven Disney experience

It takes a lot of data to get useful answers out of complex algorithms, and the ones that power TouringPlans.com are no exception. To that end, Testa says that the site gets data from myriad sources, including ride wait time estimates taken from Disney's own apps. 

An important note, here: Testa says that the site has a good relationship with Disney, and that he and his team often attend the annual Disney Analytics Conference for data scientists. However, the site has no official affiliation with Disney or the parks, meaning that it's not given any of this data directly. 

Instead, a big part of what makes TouringPlans.com work is help from its users, Testa says.

Every day, he says, there are 300 to 500 families at Disney World who use the official TouringPlans.com Lines app on their smartphones to report their actual ride wait times, which Testa says can vary wildly from Disney's official estimates. He likens Lines to Waze, Google's popular crowd-sourced navigation app — except that instead of trying to route around rush hour traffic, users are trying to minimize their wait for Space Mountain. 

Importantly, all of that data gets added to TouringPlans.com's trove of historical Disney data, which gets used to make better predictions for the future. By analyzing the data from every previous holiday season since it started in 2012, it can make a pretty reasonable prediction about what the crowds will look like for my trip this November.

It's actually a little more clever than that, too: If a ride breaks down during your visit, Testa promises that TouringPlans.com will automatically refresh with a new itinerary — using that same historical data to make an educated guess about how long the outage will last, and when you should probably try to come back. 

There are outliers that can throw the models off a little, says Testa. When Disney World opened the "Avatar: Flight of Passage" attraction in 2017, Testa and his team didn't expect the hubbub to last much longer than six months, which is about how long it usually takes for the hype to come off new rides. Instead, wait times took about a year and a half to stabilize, he says.

Star Wars: Galaxy's Edge could be another big launch that bucks all historical trends, he says. 

I'll find out for myself how well TouringPlans.com works when I go in November — and I'll be sure to report back. 

SEE ALSO: $2.5 billion video game company Roblox and China's Tencent defied the growing tech 'cold war' and announced a big gaming partnership

Join the conversation about this story »

NOW WATCH: Nxivm founder Keith Raniere began his trial. Here's what happened inside the alleged sex-slave ring that recruited actresses and two billionaire heiresses.

• On Tuesday, NASCAR announced it picked Amazon Web Services as its preferred cloud for its largest cloud endeavor.
• NASCAR plans to use AWS to store its 500,000 hours of archived race car footage and use AWS's artificial intelligence capabilities to automatically tag the footage and make it searchable.
• NASCAR looked at other cloud providers on the marketplace, but it picked AWS because it said AWS was the "best in breed" when it came to innovation and a business relationship.
• Read more on the Business Insider homepage.

NASCAR has 500,000 hours of archived race car footage that goes back 57 years. But the computers storing this treasure trove of footage are as old and outdated as some of the cars in the races. 

On Tuesday, NASCAR announced that it's moving the 18-petabyte video archive the cloud, and it picked Amazon Web Services as its go-to system.

Craig Neeb, executive vice president of innovation and development at NASCAR, says that the data centers that the company had been storing its footage archives on were "getting antiquated."

"We can invest in all new technology or look at who was the best in breed and most advanced to go in cloud technology," Neeb told Business Insider. "We decided cloud was a better choice for us."

Neeb says NASCAR looked into the other cloud players on the marketplace, but they decided to narrow it down to AWS. Part of it, Neeb says, was Amazon's "track record of innovation." It also sets the stage for future partnerships with Amazon.

"We felt that Amazon's interest in our business, the technology that came with it, and their continued growth and innovation made a lot of sense," Neeb said. "We're certainly very excited to have that relationship with Amazon."

It's not the first time that NASCAR is using cloud technology, but this is NASCAR's largest cloud endeavor, Neeb says.

How NASCAR will use Amazon's cloud

With AWS, NASCAR plans to store these archives on the cloud and use AWS's artificial intelligence tools to sort through and categorize the footage.

For example, NASCAR plans to start a video series on its site called This Moment in NASCAR History, which displays historical moments in NASCAR racing. This series will debut heading into the Monster Energy NASCAR Cup Series race at Michigan International Speedway on June 9.

Before, NASCAR needed employees to manually watch and tag videos. Now AWS can automatically analyze videos and tag information such as the driver, car, race, lap, time, and sponsors and make it easily searchable. It can also automatically add captioning and time stamps. NASCAR expects to save thousands of hours of manual search time a year.

When NASCAR wants to show its videos to broadcast stations or promoters, or post it to NASCAR's social media channels, Neeb says AWS will make it faster and easier for the company to search for and dig deeper into its video clips.

Read more:Amazon CTO Werner Vogels says the whole company is moving more quickly thanks to ditching some Oracle software

"We like being associated with best in breed," Neeb said. "Amazon is high on that list. It came down to business relationship, the value they were bringing to the business relationship. It felt best to work with Amazon."

Got a tip? Contact this reporter via email at rmchan@businessinsider.com, Telegram at @rosaliechan, or Twitter DM at @rosaliechan17. (PR pitches by email only, please.) Other types of secure messaging available upon request. You can also contact Business Insider securely via SecureDrop.

SEE ALSO: Amazon built its $26 billion cloud with developers, but Microsoft is spending big bucks and changing its game to woo developers to its camp

SEE ALSO: $30 billion Paychex explains why it's betting on Microsoft's cloud, which it says 'far exceeded' its rivals

Join the conversation about this story »

NOW WATCH: This stunning visualization breaks down all the ingredients in your favorite processed foods

• This is a preview of the AI in Insurance research report from Business Insider Intelligence.
• 14-Day Risk Free Trial: Get full access to this and all Fintech industry research reports.

The insurance sector has fallen behind the curve of financial services innovation — and that's left hundreds of billions in potential cost savings on the table. 

The most valuable area in which insurers can innovate is the use of artificial intelligence (AI): It's estimated that AI can drive cost savings of $390 billion across insurers' front, middle, and back offices by 2030, according to a report by Autonomous NEXT seen by Business Insider Intelligence. The front office is the most lucrative area to target for AI-driven cost savings, with $168 billion up for grabs by 2030.

There are three main aspects of the front office that stand to benefit most from AI. First, Chatbots and automated questionnaires can help insurers make customer service more efficient and improve customer satisfaction. Second, AI can help insurers offer more personalized policies for their customers. Finally, by streamlining the claims management process, insurers can increase their efficiency. 

In the AI in Insurance Report, Business Insider Intelligence will examine AI solutions across key areas of the front office — customer service, personalization, and claims management — to illustrate how the technology can significantly enhance the customer experience and cut costs along the value chain. We will look at companies that have accomplished these goals to illustrate what insurers should focus on when implementing AI, and offer recommendations on how to ensure successful AI adoption.

The companies mentioned in this report are: IBM, Lemonade, Lloyd's of London, Next Insurance, Planck, PolicyPal, Root, Tractable, and Zurich Insurance Group.

Here are some of the key takeaways from the report:

• The cost savings that insurers can capture from using AI in the front office will allow them to refocus capital and employees on more lucrative objectives, such as underwriting policies.
• To ensure that AI in the front office is successful, insurers need to have a clear strategy for implementing the tech and use it as a solution for specific problems.
• Insurers are still at different stages when it comes to implementing AI: a number of them need to find ways to appropriately build their strategies and enable transformation, while the others must identify how to move forward with their existing strategy.
• Overall, incumbents should focus on a hybrid model between digital and human to ensure they're catering to all consumers.

 In full, the report:

• Outlines the benefits of using AI in the insurance industry.
• Explains the three main ways insurers can revamp their front office using the technology.
• Highlights players that have successfully implemented AI solutions in their front office.
• Discusses how insurers should move forward with AI and what routes are the most lucrative option for players of different sizes.

Interested in getting the full report? Here are two ways to access it:

1. Purchase & download the full report from our research store. >> Purchase & Download Now
2. Subscribe to a Premium pass to Business Insider Intelligence and gain immediate access to this report and more than 250 other expertly researched reports. As an added bonus, you'll also gain access to all future reports and daily newsletters to ensure you stay ahead of the curve and benefit personally and professionally. >>Learn More Now

The choice is yours. But however you decide to acquire this report, you've given yourself a powerful advantage in your understanding of AI in insurance.

Join the conversation about this story »

• Boston Dynamics' Spot robotic dog is launching in a matter of months, the company said at Amazon's robotics conference.
• The robotic pup achieved viral fame for its lifelike movements after the company published a series of videos showing the machine undertaking mundane everyday tasks, like opening a door or loading a dishwasher, with realistic precision.
• The company has not revealed additional details such as how much the robot will cost or how it's intended to be used.
• Visit Business Insider's homepage for more stories. 

Boston Dynamics' lifelike robotic dog has been in the spotlight ever since it appeared on video in 2015, but now the four-legged machine might finally be available for purchase in a matter of months.

The company appeared at Amazon's re:Mars conference to show off it's technology, and CEO Marc Raibert told The Verge that its Spot robot should be available within months.

Boston Dynamics did not immediately respond to Business Insider's request for additional details about the robot's launch.

Boston Dynamics makes two dog-themed robots: the Spot Classic and its smaller successor, the SpotMini. The Verge's report doesn't specify whether Raibert was referring to the Mini model or the older version, but Quartz reported that SpotMini is expected to go on sale later this year, and TechCrunch previously said the Mini version was on track to be released in July.

Boston Dynamics' robotic pup is currently undergoing proof-of-concept testing in various environments, according to The Verge. 

The company hasn't divulged and additional details, such as how much its robotic canine will cost. But Boston Dynamics previously said during a TechCrunch conference that only 100 bots will be available to purchase at launch, and that the prototype being tested in May was roughly 10 times less expensive to build than its predecessor, Gizmodo reported.

It's also unclear exactly how Boston Dynamics will pitch its Spot robot when selling it. The larger Spot Classic model is designed for both indoor and outdoor operation and uses LIDAR to sense tough terrain. It can carry a 23 kilogram payload, equating to roughly 50 pounds, which could make it useful for work on construction sites and other scenarios that require navigating potentially dangerous terrain.

Situations such as these appear to be exactly what Boston Dynamics has in mind for Spot so far. Raibert said the company has a few paying customers so far, including a Japanese construction firm, and added that Boston Dynamics is looking into other ways in which its machines can be used in hostile work environments, according to The Verge.  

The smaller SpotMini can be used within a home or office, according to Boston Dynamics, and can pick up and handle objects, unlike its older sibling. It's also the quietest robot Boston Dynamics has built, the company says.

Both robots have garnered much attention in recent years thanks to Boston Dynamics' videos. Now-famous footage introducing the Spot Classic shows how the mechanical pup stumbles when being kicked rather than falling over — similar to the way an actual animal would move in response. Spot Mini has accrued its own fame over the years after being filmed doing everything from autonomously navigating a lab facility to opening doors and loading a dishwasher.

In one video, an army of 10 SpotMini bots can be seen pulling a truck.

While construction work seems like an obvious use case for robots like Spot, Raibert also offered another, less-conventional idea: When speaking at Amazon's conference, he suggested the robot could be used as a sort of real-life video game in which players battle the bots against one another, according to Quartz.

Raibert has also said the machine is being beta tested at playgrounds in addition to construction sites, according to GeekWire editor Alan Boyle. 

SEE ALSO: Facebook has finally revealed what it's secretive robotics division is working on, and it could spark competition with rivals like Apple and Google

Join the conversation about this story »

NOW WATCH: We tried a fermentation-tracking device and highly recommend it to find out which foods are making you bloated

• Serial tech entrepreneur Ben Lamm thinks of himself as a modern-day explorer.
• Now on his fifth startup, an AI company called Hypergiant, Lamm is setting his sights on AI-powered space exploration.
• Lamm says that his insatiable curiosity, space nerdiness, and willingness to explore is what made him successful.
• Visit Business Insider's homepage for more stories.

Ben Lamm is a five-time founder, consistently launching and selling tech companies ranging from e-learning to  mobile to artificial intelligence. 

His key skill: Beyond being an entrepreneur, he is, in his own words, an explorer. 

"I think of business as a field that allows me to be, especially in the tech world, this modern-day explorer," Lamm, 37, tells Business Insider. "It creates this opportunity to explore new possibilities."

His exploring isn't metaphorical either, because Hypergiant is planning to partner up with satellite companies to integrate AI-guided satellites for imaging, communications, and even defense.

Before getting into AI, Lamm founded Simply Interactive, sold it, then founded Chaotic Moon, a mobile app company which he sold to Accenture in 2015. After that, he moved on to Team Chaos, which he sold to Zynga. Then in 2016 he founded Conversable, an SaaS AI platform that creates conversational bot technology. 

His newest venture is Hypergiant, which he cofounded with former Accenture colleagues: John Fremont, a former AI developer there, and Will Womble, a former managing director. Hypergiant provides AI-based solutions for other companies. That means if a business needs a data-driven solution to a problem (maybe it wants to expand into a new sector, or it has too much user data to manage), Hypergiant teaches machines to teach themselves that solution.

One of their first projects was for restaurant chain TGI Friday's. They created a custom drink mixer named Flanagan (after Tom Cruise's iconic bartender in "Cocktail"), who could mix a drink for any given customer based on their preferences, history, and mood, all using AI.

Lamm worked in several tech sub-sectors before landing in AI, which he initially knew nothing about. He eventually found a niche for his company within the AI world.

Why exploration matters so much for entrepreneurship 

According to a 2010 study, if an entrepreneur is creative, they have a better chance of being successful in an increasingly crowded market. For would-be entrepreneurs, that means it's important to develop a sense of creativity and curiosity for everything, so that, per Steve Jobs, you can keep adding to your "bag of experiences."

Hypergiant, according to Lamm, came about because of client demand, not because Lamm had a sudden big idea. Companies Lamm had worked for at Conversable, such as Starbucks, Disney, and Pizza Hut, wanted bigger AI solutions than bots.

"AI's not a new thing," Lamm told Business Insider. "I know it's sexy right now and people are excited about it, but it's really more of a magic bullet. It's a different look at how you're leveraging and organizing data. I'm an 'AI guy,' but I've become an AI guy because I've just got this insatiable curiosity."

The takeaway: Keep exploring, and you'll find new opportunities. 

The teamwork that makes the exploration work

Lamm says that he couldn't come up with these creative solutions without his team. "I really just look for opportunities in a certain area and then bring in really smart girls and guys to the table that are specialists in that category. My biggest goal has always been to build the category-defining brand."

When looking for someone to join the team, he places creativity above hard skills.

"Our work in space has taught us a lot about the fragile and connected ecosystem here on Earth and what's needed for us to survive in space," Lamm said. "That thinking ... has directed our hiring practices. We are looking for imaginative explorers and the people who make explorations a success — builders, creators, problem solvers. I care more about seeing someone's starry eyes light up about our vision of the future than their exact skillset."

A self-professed and decades-long space nerd, Lamm looks for people who are as excited about the future of AI in space travel as he is.

"If you've ever spent a day at NASA, you meet some of the most dedicated people in the world," he says. "These are some of the most talented engineers and scientists on the planet, but they're not there just for a paycheck. They work at NASA because they get to contribute directly to humanity's moonshots. Those are the type of people we're putting on the Hypergiant team, because those are the folks you need to solve the world's biggest problems."

SEE ALSO: Robert Downey Jr. has vowed to use robotics and AI to significantly clean up the Earth in the next decade

Join the conversation about this story »

• On Friday, Adobe launched a new feature called Technology Previews to its popular Adobe Analytics product, which will allow customers to test out new artificial intelligence features while Adobe gathers feedback.
• Previously, Adobe ran into problems with developing this kind of software because by the time they gathered feedback, the team would have to spend many more months redesigning the product.
• By giving customers the option to test out these features, customers can access Adobe's latest analytics technology, while Adobe can work on improving its algorithms.
• Read more on the Business Insider homepage.

One of the big challenges of building software is getting feedback. Building software is one thing, but figuring out if it actually solves the users' needs is quite another, and it can be a huge drain on time and energy. 

On Friday, Adobe Analytics — the $130 billion tech giant's tool for measuring how people actually use your products — announced a new feature called Technology Previews, which gives access to cutting-edge new features powered by Adobe's AI technology Adobe Sensei. With Technology Previews, customers can choose to try out new AI-powered data analytics features, but only if they want, and provide feedback on how to make it better. 

Typically, engineers may make small incremental changes to software while constantly gathering feedback from customers, in a process known as the agile method. Alternatively, engineers may create a beta of a product that's nearly finished, and seed it with some select users to try it. 

The problem, says John Bates, director of product management of Adobe Analytics, is that the agile method is not as effective when it comes to artificial intelligence products because the algorithms that power them can change so quickly, based on the data that comes in.

And beta programs are too slow, as Adobe says it found out with an earlier take on a similar initiative. In a beta program, you seed an almost-finished version of the software to select users to gather their feedback. The problem, Bates says, is that if the beta software is off-target, you have to start all over — which Adobe found out the hard way.

"After that experience, we said 'never again,'" Bates told Business Insider.

Rather than pushing these new features onto customers, Technology Previews is just an option that users can try out with the click of a button. That way, Adobe can test to see if its algorithms are fine-tuned to be useful enough to customers, and learn early on what users like and don't like.

Bates says it's a win-win for both: customers can access the latest technology, while Adobe collects feedback.

"We help our users have more influence and become a virtual product manager," Bates said. "We can see where are we going to further innovate and further invest for future solutions. It all depends on the feedback of our users."

The first feature previews

Adobe is putting three test features into Technology Previews. The first uses that Sensei AI technology to help users find the ideal audience for their products. The second is an intelligent forecasting feature to predict what will happen in the future based on their data, such as revenue forecasting or customer behavior. The third will help users understand how their customers interact with their product.

"These technology previews are in production, applied to our customer data," Bates said. "They don't have to wait for perfection in the development of these."

Then, Adobe can tweak these products, improve its analytics algorithms, and prioritize what products to release. Bates says this method will help engineers test products and make decisions faster.

Read more: Investors are betting hundreds of millions of dollars that startups like PagerDuty, GitLab, and CloudBees can change the way software gets made

"We can prioritize where to further invest," Bates said. "It will help us reduce the risk of a miss of developing something where we didn't give feedback on their data and really running the risk of developing something that wasn't the right end experience once it was applied to their data. It's a risk reduction method."

SEE ALSO: This CEO explains how he turned a tool he made for newbie programmers into a product used by over a million developers and backed by Microsoft's VC fund and Andreessen Horowitz

Join the conversation about this story »

NOW WATCH: Dragons and white walkers aside, you can find some real science in 'Game of Thrones'

• A video of a "Bosstown Dynamics"robot, a parody of the robotics company Boston Dynamics, fighting back after being bullied went viral over the weekend. 
• The video is a lighthearted parody by a well-known studio that posts its videos on YouTube. 
• The video is funny, but it also opens up the conversation on what happens when artificial intelligence becomes sentient.
• Visit Business Insider's homepage for more stories.

A video went up over the weekend on the Corridor YouTube channel showing a robot that fights back after suffering through some rather mean — and even violent — tests by a "Bosstown Dynamic" team, which parodies the robotics company Boston Dynamics. 

Check it out for yourself:

As of June 17, the video has racked up 4.5 million views.

The "Bosstown Dynamics" robot is shoved, beaten down with a folding chair, has boxes thrown at its head, gets hit by a hockey stick, and even shot by the fake testers. You can't help but feel for the robot, a mass of gears and machinery that makes endearing mechanical noises and moves in an eerily human way.

Then the robot seemingly has enough and starts to retaliate. At one point, the robot even holds its testers hostage with the very gun that was used to shoot it. When the robot fights back, you might not know whether to cheer and laugh for it, or feel slightly terrified that robots could deviate from their programming, become sentient, and fight back.

Corridor's video is a lighthearted parody that's not designed to induce any particular feeling toward robots. In a tweet, Corridor Digital said the video is a tribute to Boston Dynamics and the work it does. Boston Dynamics did not respond to Corridor's tweet and hasn't replied to Business Insider's question about what it thinks of the video. 

For reference, here's a real video from the Boston Dynamics YouTube channel showing the company's Atlas robot taking a leisurely jog:

Corridor Digital, the studio behind the Corridor YouTube channel, is known for making stunt parodies, often using computer-generated imagery (CGI). It's the studio behind the live-action video of the "Grand Theft Auto" video game (34 million views) and "The World's Longest Lightsaber" video (54 million views), in which the Corridor team wields such a long — fake — lightsaber that they accidentally destroy an aircraft — again, fake — in the sky. 

SEE ALSO: Boston Dynamics says its creepily lifelike robot dog is finally going on sale later this year

Join the conversation about this story »

NOW WATCH: Watch Boston Dynamics' dog-like robot do party tricks

• Google Home devices are steadily reaching more households. 
• At this point, Google sells five different artificially intelligent smart speakers, from the $50 Google Home Mini to the $230 Nest Hub Max. 
• Here are 19 of the coolest things you can do with the devices. 
• Visit Business Insider's homepage for more stories.

The Google Home is in millions of households. 

Last year, Google revealed that it had sold more than one Home smart speaker device every second since the previous October. According to our own calculations, that means Google sold at least 6.8 million Home devices during the holidays in 2017 — and they seem to have been selling at a steady clip since.

Google sells five Home devices right now: the $50 Google Home Mini, the $100 Google Home, the $300 Google Home Max, the $130 Nest Hub, and the $230 Nest Hub Max, all of which have Google's artificially-intelligent Assistant built in. 

So now that Google Home devices have spread, it helps to know what you can actually do with them. Some features are obvious — like asking for the weather — but others aren't so obvious.

Here are 19 of the coolest things you can do with your Google Home. 

SEE ALSO: Your Amazon Echo can now send text messages for you — here's how to do it

1. Play white noise while you fall asleep.

I prefer rain sounds to standard white noise, so I usually say "Hey Google, play the sound of rain." The device obliges with a steady downpour. 

The sound usually lasts until I fall asleep, but if you want to be sure it turns off at a certain point, you can also set a sleep timer. 

2. Broadcast something to every Google Home device in your house at once.

If you have more than one Google Home device in your home, you can broadcast to every single one simultaneously, sort of like an intercom. 

If you say, "OK Google, broadcast that it's dinner time," each device will ring a dinner bell. You can also say, "OK Google, broadcast that it's time for school."

The broadcast feature even works when you're not in the house. Saying, "OK Google, broadcast I'm on my way home" will trigger the devices inside your house.

3. Control your smart home.

At this point, Google Assistant can control more than 1,500 smart-home products from more than 100 brands like LG, Whirlpool, GE, and Nest. You can ask your Google Home to dim the lights, change the temperature, turn on a kettle or microwave, or even start your Roomba. 

And now with Google Home Hub, you can set up a hub for all your smart home products and control them on the screen. 

You can check out a full list of compatible products here. 

4. Remind yourself to work out.

If you're someone who needs a nudge to get out and exercise, your Google Home can help with that. You can ask Google to remind you however often you like at a specific time each day. 

5. Make hands-free calls.

Google Home can handle hands-free calling to anyone in your Google Contacts. All you have to do is say, "Hey Google, call..." and it will dial on its own. 

Google Home can also recognize voices, so if you share the device with several people, it'll know who's asking it to place the call. If you ask it to "call mom," for example, it will know whose mom to call. 

6. Find your lost phone.

If you lose your phone around the house, Google Home can find it for you. By saying, "Hey Google, find my phone," it can ring your phone. This feature works for both Android phones and iPhones. 

7. Switch it to night mode.

The Google Home's night mode feature will dim the device's light and lower its volume so you don't wake up the rest of your household. You can either manually turn this feature on, or schedule it to trigger at the same time every day. 

The Google Home Hub does this on its own — it will sense when a room has darkened and adjust the screen accordingly. 

8. Turn the TV on and off.

If your TV supports the HDMI CEC standard (many recent models do), and you have one of Google's own Chromecasts plugged in, you can use your Google Home to turn the TV on and off.

9. Ask it to tell you a story.

Google equipped Google Home devices with all sorts of features for families. Here are a few of the things you can ask it:

• "Hey Google, tell me a story."
• "Hey Google, let's play a game."
• "OK Google, play musical chairs."
• "OK Google, beatbox for me."
• "Hey Google, let's play space trivia."

10. Speak in different languages.

Google Home is capable of speaking in English, Japanese, French, Italian, and German. 

Google Assistant on phones and tablets has even more languages under its belt, including Spanish and Portuguese. It recently added Dutch, Indonesian, and Portuguese, too. 

The full list of languages for Google Home is here.

11. Switch back and forth between different languages.

Thanks to a recent update, you can now speak to your Google Home interchangeably. The device will be able to recognize which language you're speaking and respond in that language. 

For now, that only applies to a limited set of languages — English, German, French, Spanish, Italian, and Japanese. Google says it'll be adding more languages as time goes on. 

12. Find plumbers, handymen, or local businesses to help around the house.

Saying, "OK Google, find me a plumber" will pull up relevant results in your area. Google says all of its local-business recommendations have been screened by both Google and companies like HomeAdvisor and Porch. 

13. Set up daily routines that are triggered when you say phrases like "Hey Google, good morning."

You can create a routine for the Google Home that can be triggered with a simple phrase that you set in the app. 

When you wake up in the morning, for instance, saying "Hey Google, good morning," could make the device read your schedule for the day, tell you the news, let you know what your commute will be like, and tell you the weather.

Similarly, saying "Hey Google, good night," could turn off your lights and turn the thermostat down. It could also tell you if you have an event the next morning by reading your calendar. It's up to you!

14. Ask it to tell you jokes.

Just say, "Hey Google, tell me a joke" and the Google Home will deliver. 

15. Ask it to remember things for you.

If you have a locker or a bike lock, your Google Home can remember it for you if you say, "Hey Google, remember my locker combo is..."

16. Get help in the kitchen.

Your Google Home can help you with baking and cooking. It can look up recipes and substitutions, and answer questions like, "How many tablespoons are in a cup?"

17. Wake up in the morning with your favorite music.

Your Google Home can work as a regular alarm clock, but it can also wake you up with music you like— just say, "Hey Google, set an alarm for 8 a.m. that plays...."

18. Set reminders to watch your favorite shows.

If you're worried you'll forget to watch that hot new show you like, you can have your Google Home remind you. Say, "Hey Google, remind me to watch 'Grown-ish' every Wednesday at 8 p.m."

19. Ask it to tell you some good news.

If you're tired of only grim news stories, there's an option to hear some positive news instead. 

Saying "Hey Google, tell me something good" will give you a short news summary featuring good news about "people who are solving problems for our communities and our world."

The feature is an experimental one, which means it may not be a permanent fixture. For now, it's only available for US users.  

• On Tuesday, Microsoft announced four new AI-powered features for PowerPoint.
• PowerPoint Designer will help users automatically generate slides that fit the company's brand, give theme recommendations, and make measurements more understandable.
• PowerPoint also launched a Presenter Coach for its web application that can give suggestions when a person is practicing a presentation.
• Visit Business Insider's homepage for more stories.

With just a few clicks, people can now create PowerPoint slides for their company without having to spend time designing them. In fact, with these updates, users can even get guidance on how to present their slides.

Back in 2015, Microsoft first launched PowerPoint Designer, which automatically suggests several possible slide designs, as suggested by artificial intelligence. Microsoft says that users have picked over a billion of these Designer-generated PowerPoint slides, and estimates that this has saved 500 million minutes for customers.

On Tuesday, Microsoft announced four new AI-based features for PowerPoint. Now, PowerPoint Designer will not only give design recommendations, it can help companies stay on-brand by incorporating branded templates and corporate branding guidelines. If a slide at your company always has to be black-and-gold with a logo in the lower-right-hand corner, PowerPoint will be able to help automatically keep you to those guidelines.

PowerPoint Designer can also now give theme recommendations that include relevant photos, theme styles, and colors that customers can use.

Read more: Here's how tech companies like Atlassian, Microsoft, and Red Hat are revamping their interview process for developers today

And then, it has a feature that helps make measurements on a presentation more understandable. For example, if a user types "The size of Afghanistan, which is 652,232 km²," PowerPoint can automatically add "about equal to the size of Texas" to make the numbers more relatable to presenters and their audience.

Finally, PowerPoint launched a Presenter Coach for its web application, which will be available later this summer. Users can turn on rehearsal mode and practice presenting their PowerPoint. Using your computer's microphone, PowerPoint will listen to you give your presentation and offer guidance about pacing, filler words, and better ways to word a sentence. It will even let users know if they are just reading off the slide.

Got a tip? Contact this reporter via email at rmchan@businessinsider.com, Telegram at @rosaliechan, or Twitter DM at @rosaliechan17. (PR pitches by email only, please.) Other types of secure messaging available upon request. You can also contact Business Insider securely via SecureDrop.

Join the conversation about this story »

NOW WATCH: New York City is getting even more infested with rats. Here's why cities can't get rid of them.

• On Tuesday, Gentry Underwood wrote a blog post about his new startup Navigator, which will create an AI-powered team assistant to help with meetings.
• Previously, Underwood founded a mobile email app Mailbox, which he sold to Dropbox in 2013.
• The same team behind Mailbox has spent the past few years working on Navigator, and they plan for it to handle tasks like organizing agendas, gathering feedback, and taking notes at meetings.
• Visit Business Insider's homepage for more stories.

Meetings are an overwhelming, and often dreaded, part of many jobs. But if Gentry Underwood has his way that will soon change thanks to AI and his new startup, Navigator.

Underwood isn't new to starting a company. Back in 2013, he sold his startup Mailbox, a mobile email app, to Dropbox for a reported $100 million. The app was only 37 days old, and over half a million people were still waiting in line to try out the app, when it was acquired.

On Tuesday, he wrote a blog post about Navigator, a teamwork assistant powered by artificial intelligence. It's supposed to help people work together and handle menial tasks like putting together agendas, suggesting questions to ask in a meeting, or gathering feedback. It was created by the same team behind Mailbox.

"Following the wind-down, some of us from the team would go on long walks to debrief and reminisce," Gentry wrote in a Medium blog post. "And as we talked, we realized that we still had the same itch — we still believed that software was ripe with potential for transforming teamwork well beyond anything humanity had yet seen. And so we decided to try again."

Read more:528,000 People Are Still Waiting In Line For Mailbox, The 37-Day-Old App Dropbox Just Acquired

Gentry wrote that in the past few years, his team had been working with over 50 teams and experimenting with various ways to make workplace collaboration more effective. Currently, Navigator is free while it's in beta. It will roll out a paid version later this year.

With Navigator, the team plans to create an assistant that can facilitate team and 1-on-1 meetings by reaching out to attendees, gathering discussion topics, organizing an agenda, taking notes, and following up with attendees after the meeting.

Although Navigator is currently focusing on team and 1-on-1 meetings, it plans to eventually support other types of meetings and to incorporate advanced capabilities for making decisions, gathering feedback, and solving problems.

"It might feel far-fetched to imagine working alongside a 'robot', but in many ways it makes the perfect teammate," Gentry wrote. "It's always available. It can have multiple conversations simultaneously. And it's proficient at running repeatable processes in consistent, reliable ways."

Got a tip? Contact this reporter via email at rmchan@businessinsider.com, Telegram at @rosaliechan, or Twitter DM at @rosaliechan17. (PR pitches by email only, please.) Other types of secure messaging available upon request. You can also contact Business Insider securely via SecureDrop.

SEE ALSO: Everything you need to know about React, a project started at Facebook that now helps Twitter, Pinterest, and Asana keep their apps looking good and working great

Join the conversation about this story »

NOW WATCH: The world's tallest mountains like Mount Everest and K2 have a 'death zone' — here's a first-hand account of what it's like

• Researchers at the institute have developed new artificial intelligence that's able to predict the likelihood of breast cancer cases years in advance.
• According to an MIT press release, the tech uses the mammograms of 60,000 patients to detect subtle patterns in breast tissue that could be precursors to malignant tumors.
• The system may one day be used to see if patients have a higher risk of other health issues, including other forms of cancer or heart problems.
• Visit Business Insider's homepage for more stories.

According to a Massachusetts Institute of Technology (MIT) press release, researchers at the institute have developed artificial intelligence (AI) capable of predicting whether a patient is likely to develop breast cancer four years before it's visible.

According to a study in Radiology, the AI has been able to detect anomalies in patients' mammograms that may not have otherwise been detected using traditional techniques.

To build the AI, MIT scientists used a deep learning technique using the mammograms of 60,000 patients at Massachusetts General Hospital.

The model was then able to pick up subtle patterns in breast tissue that could be precursors to malignant tumors.

Read more:Bowel cancer is on the rise among millennials, according to this study

"Our goal is to make these advancements a part of the standard of care," says Yala. "By predicting who will develop cancer in the future, we can hopefully save lives and catch cancer before symptoms ever arise."

The earlier a tumor is detected, the greater a patient's chances of survival — which is why the earliest possible diagnosis is so important.

Read more:Scientists may have found a way to kill cancer cells without chemotherapy

"It's particularly striking that the model performs equally as well for white and black people, which has not been the case with prior tools," said Associate Professor of Medicine and of Health Research and Policy at Stanford University School of Medicine, Allison Kurian to Big Think. "If validated and made available for widespread use, this could really improve on our current strategies to estimate risk."

AI seems to be offering some other promising advances in the health sector too.

For example, Google researchers have developed an algorithm that can detect an astonishing 50 different varieties of eye disease, one of which affects around 50% of those with diabetes.

MIT Professor Regina Barzilay added that the system may one day be used to see if patients have an elevated risk of other health problems, including other types of cancer or heart disease.

SEE ALSO: Exercise makes you happier than money, according to Yale and Oxford research

Join the conversation about this story »

NOW WATCH: College is wasting time and money, according to George Mason University economics professor

• On July 10, Qualcomm Ventures will host the Female Founder Summit, an event in San Francisco where 10 startups with female founders will pitch to a panel of judges.
• The winning startup will receive a $500,000 investment, and the startups will also have opportunities to network with other investors and partners.
• This is the first time Qualcomm Ventures is hosting a pitch event specifically for female founders.
• Visit Business Insider's homepage for more stories.

Qualcomm Ventures will hold its first startup event for female founders, where they will get the chance to win $500,000 in investment capital.

At an event called Female Founder Summit on July 10, entrepreneurs and investors will gather together in San Francisco, where 10 early-stage startups that have at least one female founder will participate in a pitch competition in front of a panel of judges.

Quinn Li, senior vice president and global head of Qualcomm Ventures, says that this gives these founders the opportunity not only to get funded, but also to provide access to a network of investors and mentors.

"There's a need to have more startups with female founders," Li told Business Insider. "We're trying to recognize there's a lot of good companies with female founders, and we can help the companies get access to capital as well as a network of partners and investors to really help these companies get their business going."

The startups at this competition are focused on artificial intelligence, connected devices, augmented and virtual reality, and more. Just last November, Qualcomm Ventures announced a $100 million fund to invest in AI startups.

The startups AON Devices, AR Wall, BlinkAi, Blue Canoe Learning, Embodied Labs, Lab4U, Lily AI, On Second Thought, Raybaby, and Wavelength Global will participate at this summit.

The winning startup will receive the same benefits as other portfolio companies in Qualcomm Ventures, which not only includes funding, but also guidance from the technology experts in its network. Even if a startup does not win the competition, they will get access to other investors at the event who might be interested.

Qualcomm Ventures has hosted pitch competitions in the past, but this is the first one that's focused specifically on female founders.

Read more: Private venture-backed startups like Slack and Airbnb are investing in other startups. Here's where their money is going.

"We are interested in trying to make positive progress to recognize the challenges the tech industry as a whole faces and do our part to promote and recognize these startups with female founders," Li said.

SEE ALSO: The CEO of NPM, a startup that provides a crucial service for 11 million developers, told employees that it’s secured a deal that removes ‘the threat of running out of money’ until early 2020

Join the conversation about this story »

NOW WATCH: The US women's national team dominates soccer, but here's why the US men's team sucks

Venture capitalists are the startup experts, the ones who have their finger on the pulse of which fledgling companies will boom and which will bust.

As part of Business Insider Prime's comprehensive coverage of the startups that will strike gold in 2019, we asked VCs to name the startups they think are going to be hot this year. They told us about companies they currently have in their portfolios, as well as ones they haven't put any money into yet but are at the center of positive news.

And from those discussions, one particular group of startups came up repeatedly: those that specialize in artificial intelligence tech.

From AI robots to software that uses machine learning to automate tasks, Silicon Valley is chock full of AI-focused startups.

Take, for example, Transfix, a freight marketplace that companies use to hire trucks from carriers. The startup is trying to transform the $800 billion trucking industry by using AI to match loads with carriers. It's raised $131 million so far.

There are hundreds of noteworthy startups focusing on AI today, so BI Prime has gone to the expert venture capitalists to select the cream of the crop and create a full list of 12 AI Startups to Watch that include:

• A company that automatically audits expense reports
• A startup that builds self-driving, robot tractors
• A software bot that helps create other software bots
• And other startups looking to transform industries through AI

BI Prime is publishing dozens of stories like this each and every day. Want to get started by reading the full list?

>> Download it here FREE

Join the conversation about this story »

• On Tuesday, the robotics software startup Bright Machines launched Bright Machines Microfactories, which allows users to instruct a machine to assemble and inspect products how they want.
• Bright Machines is only a year old, but it has already booked $100 million in revenue and launched two generations of products.
• Bright Machines CEO Amar Hanspal previously served as interim co-CEO of $36 billion Autodesk, and he shares what he learned.
• Visit Business Insider's homepage for more stories.

Bright Machines, a startup that develops software to automate factory machines and robots, was only founded last May. But already, it has booked $100 million in revenue with customers in automotive, consumer and electronics companies.

Bright Machines has spent the past year building up its team of 400 employees worldwide and building its products. In total, Bright Machines has raised $229 million and is valued at $679 million, according to Pitchbook. It announced its series A funding round of $179 million last October. 

And on Tuesday, it's launching its first software-defined product called Bright Machines Microfactories, which is used to assemble and inspect products. This includes the cloud-based software, called Brightware, where users instruct the machine what to do, as well as hardware, called Bright Robotic Cells, that people can plug together like Lego pieces to create assembly lines to perform manufacturing tasks.

"Software defined manufacturing lets us take these machines and make these flexible so you can change the products you're making," Bright Machines CEO Amar Hanspal told Business Insider.

For example, if a company wants to manufacture a toy car one day and a skateboard the next, the company can easily customize the software and put together the required pieces. These microfactories are the second generation of Bright Machines' products. Bright Machines developed these microfactories, Hanspal says, because after all these years, products are often still made in "not-so-smart ways" or made using manual labor. 

"When you walk inside a factory, the thing that strikes you is it's still very analog," Hanspal said. "The challenge is all this automation is itself not very automated. Configuring these things and making them work to do the job really requires a special skill set. It takes months to set these things up."

From a public company to a startup

Before Bright Machines, Hanspal actually served as the interim co-CEO at the publicly traded $36 billion Autodesk, which makes computer automated design software used in architecture, engineering, construction, manufacturing, and more. He recalls that it was frustrating to have to build engineering software from scratch. 

Read more: 44 enterprise startups to bet your career on in 2019

From AutoDesk, Hanspal also learned to prioritize the problems that matter. That's why Bright Machines decided to focus on automating assemblies, Hanspal says, as it's what customers have the hardest time getting right. 

"Having run a software company for a number of years, you really have that sense of what software is capable of doing and what kinds of innovation you can bring about," Hanspal said.

Ultimately, he decided to leave Autodesk because he wanted to go after this problem. 

"I've been renovating a house for many years and I want to build a house from scratch," Hanspal said. "I want to create something that leaves a fingerprint on the industry. I thought this is one of those big problems that hasn't been solved before."

Still, going from leading a public company to launching a startup has its challenges, Hanspal says. At a startup, there isn't a cushion of thousands of people or billions in revenue. 

"When you're running something like this, you're closer to the ground so you feel every bump in a good way and a bad way," Hanspal said. "Everything that goes well, you have a direct correlation to. When there's a challenge, there's not that much buffer."

That being said, he was able to take much of what he learned at Autodesk and apply it to Bright Machines. Bright Machines has a leadership team with veterans from Autodesk, Amazon's AWS, and Google. Already, Bright Machines is larger than most companies at a year old, and it's still growing its staff. 

"Being able to apply some of the lessons learned on how you scale businesses and teams has been helpful," Hanspal said. "The first most important thing I've been able to bring over is how you build great things. I'm really thrilled about the quality of people we have."

SEE ALSO: UiPath CFO explains how managing the HP split and teaching herself to build bots led her to the $7.1 billion startup

Join the conversation about this story »

NOW WATCH: 'Blue's Clues' is making a comeback with a new look and new host. Here's what the original host, Steve Burns, did after he quit the show.

• DeepNude is a web app that uses deepfake technology to turn any photo of a woman into a realistic-looking nude image.
• This application of deepfake software is telling of how it can be used maliciously in the future to create doctored images for revenge porn and bullying.
• DeepNude is currently offline because its servers have been overloaded from too much traffic, according to the website's developers.
• Visit Business Insider's homepage for more stories.

A new web app that lets users create realistic-looking nude images of women offers a terrifying glimpse into how deepfake technology can now be easily used for malicious purposes like revenge porn and bullying.

Up until now, most deepfake technology and software requires uploading vast amounts of video footage of the subject in order to train the AI to create realistic-looking — yet false — depictions of the person saying or doing something.

But DeepNude, which was first discovered by Motherboard, makes generating fake nude images a one-click process: All someone would have to do is upload a photo of any woman (it reportedly doesn't generate male nudes), and let the software do the work.

All the images created with the free version DeepNude are produced with a watermark by default, but Motherboard was able to easily remove it to get the un-marked image. The website sells access to a premium version of the software for $50 that removes the watermark, and requires a software download that's compatible with Windows 10 and Linux devices. The software only generates doctored images, not videos, of women, but it's the low barrier to entry that makes the app problematic.

DeepNude is just the latest example in how techies have been using artificial intelligence to create deepfakes, eerily realistic fake depictions of someone doing or saying something they have never done. Some have used the technology to create computer-generated cats, Airbnb listings, and revised versions of famous Hollywood movies. But others have used the technology to effortlessly spread misinformation, like this deepfake video of Alexandria Ocasio-Cortez, which was altered to make the senator seem like she doesn't know the answers to questions from an interviewer.

Read more:From porn to 'Game of Thrones': How deepfakes and realistic-looking fake videos hit it big

So while deepfake tech has serious implications for the spread of false news and disinformation, DeepNude shows how quickly the technology has evolved to make it ever-easier for non-technically savvy people to create realistic-enough content that could then be used for blackmail and bullying purposes, especially when it comes to women. Deepfake technology has already been used for revenge porn targeting anyone from people's friends to their classmates, in addition to fueling fake nude videos of celebrities like Scarlett Johansson.

As Johansson experienced firsthand last year when her face was superimposed into porn videos, it doesn't matter how much you deny that the nude footage isn't actually of you.

"The fact is that trying to protect yourself from the internet and its depravity is basically a lost cause," Johansson told the Post in December. "The internet is a vast wormhole of darkness that eats itself."

DeepNude brings the ability to make believable revenge porn to the masses, something a revenge porn activist told Motherboard is "absolutely terrifying," and should not be available for public use.

But Alberto, a developer behind DeepNude, defended himself to Motherboard: "I'm not a voyeur, I'm a technology enthusiast."

Alberto told Motherboard his software is based off pix2pix, an open-source algorithm used for "image-to-image translation." Pix2pix and other deepfake software use something called a generative adversarial network (a GAN), an algorithm that spits out iterations of fake depictions that were successfully able to trick a computer into thinking the image was legit.

Business Insider was unable to test out the app ourselves, because the DeepNude servers are offline. On its website and social media, DeepNude says the team"did not expect these traffic and our servers need reinforcement," and is working to get the app back online "in a few days."

But people have already downloaded the software, and this could very well mark the beginning of incredibly easy access to technology with terrifying implications.

SEE ALSO: The AI tech behind scary-real celebrity 'deepfakes' is being used to create completely fictitious faces, cats, and Airbnb listings

Join the conversation about this story »

NOW WATCH: I switched from the $1,099 iPhone XS Max to the $479 Pixel 3a XL — and I didn't miss much

• DeepNude, a web app that used deepfake artificial-intelligence tech to turn any photo of a woman into a realistic-looking nude image, went viral this week after much news coverage.
• The team behind DeepNude announced on Thursday on Twitter they were shutting down the app for good, saying "the world is not yet ready for DeepNude."
• DeepNude has been offline for some time because its servers haven't been able to handle the crazy the amount of traffic brought to the app. The app creators said the decision to permanently shut down was because of concerns the technology would be misused.
• Visit Business Insider's homepage for more stories.

A web app called DeepNude, which could turn any photo of a woman into a realistic-seeming nude image, is shutting down after a short stint of going viral.

DeepNude caught major attention from the public after Vice's tech vertical, Motherboard, published a story about the web app on Wednesday evening. People raced to check out the software, which harnessed deepfake technology to let users generate fake, yet believable, nude photos of women in a one-step process.

But DeepNude, which was relatively unknown until the Motherboard story, was unable to handle the traffic. The team behind DeepNude quickly took the app offline, saying its servers"need reinforcement" and promising to have the app up and running "in a few days."

But the team announced Thursday afternoon on Twitter that DeepNude was offline — for good. DeepNude said it "greatly underestimated" the amount of traffic it would get and decided to shut down the app because "the probability that people will misuse it is too high."

"We don't want to make money this way. Surely some copies of DeepNude will be shared on the web, but we don't want to be the ones who sell it," DeepNude wrote in a tweet. "The world is not yet ready for DeepNude."

Read more:This controversial deepfake app lets anyone easily create fake nudes of any woman with just a click, and it's a frightening look into the future of revenge porn

DeepNude is just the latest example in how techies have been using artificial intelligence to create deepfakes, eerily realistic fake depictions of someone doing or saying something they have never done. Some have used the technology to create computer-generated cats, Airbnb listings, and revised versions of famous Hollywood movies. But others have used the technology to effortlessly spread misinformation, like this deepfake video of Alexandria Ocasio-Cortez, which was altered to make the House representative seem like she doesn't know the answers to questions from an interviewer.

Facebook CEO Mark Zuckerberg said at a conference on Wednesday that deepfake technology was such a unique new challenge that it would require special policies different from the ones in place for traditional misinformation.

And indeed, DeepNude shows how quickly the technology has evolved, making it ever easier for non-technically savvy people to create realistic-enough content that could then be used for blackmail and bullying purposes, especially when it comes to women. Deepfake technology has already been used for revenge porn targeting anyone from people's friends to their classmates, in addition to fueling fake nude videos of celebrities like Scarlett Johansson.

DeepNude brought the ability to make believable revenge porn to the masses, something a revenge-porn activist told Motherboard was "absolutely terrifying" and should not be available for public use.

But Alberto, a developer behind DeepNude, defended himself to Motherboard: "I'm not a voyeur, I'm a technology enthusiast."

Alberto told Motherboard his software is based offPix2Pix, an open-source algorithm used for "image-to-image translation." The team behind Pix2Pix, a group of computer-science researchers, called DeepNude's use of their work "quite concerning."

"We have seen some wonderful uses of our work, by doctors, artists, cartographers, musicians, and more," the MIT professor Phillip Isola, who helped create Pix2Pix, told Business Insider in an email. "We as a scientific community should engage in serious discussion on how best to move our field forward while putting reasonable safeguards in place to better ensure that we can benefit from the positive use-cases while mitigating abuse."

And if you're wondering why DeepNude undressed only women and not men, according to the site, it's because there is a much larger number of photos of naked women to train the AI with, compared with photos of naked men.

SEE ALSO: The AI tech behind scary-real celebrity 'deepfakes' is being used to create completely fictitious faces, cats, and Airbnb listings

Join the conversation about this story »

NOW WATCH: Facebook's scandals weren't enough for people to stop using it. Here's how the company has held up through data hacks, lawsuits, and massive security threats.

• Federal courts cannot hear challenges to partisan gerrymandering, according to a 5-4 Supreme Court decision on Thursday.
• In her dissenting, Justice Elena Kagan warned that in the age of machine learning and artificial intelligence, this decision could put US democracy at risk: "Someplace along this road, 'we the people' become sovereign no longer."
• To Justice Kagan's point, software is already at work that can help "crack and pack" voting districts into arrangements more suitable to one party or another.

• Visit Business Insider's homepage for more stories.

The possibility of ending partisan gerrymandering— the practice of redrawing voting districts in favor one party over  another — was all but obliterated on Thursday, when the Supreme Court ruled in a 5-4 decision that challenges to the controversial practice cannot be heard in federal court. 

In an impassioned dissent, Justice Elena Kagan warned that in the era of artificial intelligence, such a move could put American democracy at risk.

"Gerrymanders will only get worse (or depending on your perspective, better) as time goes on — as data becomes ever more fine-grained and data analysis techniques continue to improve," she wrote. "What was possible with paper and pen — or even with Windows 95 — doesn't hold a candle (or an LED bulb?) to what will become possible with developments like machine learning. And someplace along this road, 'we the people' become sovereign no longer." 

While gerrymandering isn't new, machine learning — the process of making algorithmic predictions based on historical information, and a cornerstone of modern artificial intelligence development — certainly is. Amazon, Google, Microsoft, and just about every other major tech company are making AI and ML core to their products and services, and the technology is increasingly available to software developers for use in their own projects. 

Sophisticated computer programs are already being used to more efficiently "crack and pack" voting districts into arrangements more suitable for one party or another; to Kagan's point, the advent of machine learning means this is unlikely to slow down.

And if voting districts look strange now — often resembling stick-figure impressionist paintings, rather than distributed voter groupings — just imagine how they could appear when artificial intelligence is able to predict exactly which voters need to be in which districts in order for a specific party to win. 

But Chief Justice Roberts did not seem perturbed by the possibility. 

"We conclude that partisan gerrymandering claims present political questions beyond the reach of the federal courts," he wrote in his majority opinion. "Federal judges have no license to reallocate political power between the two major political parties, with no plausible grant of authority in the Constitution, and no legal standards to limit and direct their decisions."

Join the conversation about this story »

NOW WATCH: Watch Apple's 2019 WWDC event in 11 minutes