Lifeboat Foundation

Of all the holy grails in robotics, learning may well be the holiest. In an era when the term “general purpose” is tossed around with great abandon, however, it can be difficult for non-roboticists to understand what today’s systems can — and can’t — do. The truth of it is that most robots these days are built to do one (or a couple, if you’re lucky) thing really well.

It’s a truth that spans the industry, from the lowliest robot vacuum to the most advanced industrial system. So, how do we make the transition from single to general purpose robotics? Certainly, there are going to be a lot of stops in multipurpose land along the way.

The answer is, of course, robot learning. Walk into nearly any robotics research lab these days and you will find teams working on tackling the issue. The same applies to startups and corporations, as well. Look at companies Viam and Intrinsic, which are working to lower the bar of entry for robot programming.

The image generator inside the AI-powered Bing Chat is getting a big upgrade today: Microsoft announced that OpenAI’s latest DALL-E 3 model is now available to all Bing Chat and Bing Image Creator users. It has been rolling out over the last week or so, first to Bing Enterprise users and then to Bing Image Creator, but now it’s open to everyone.

Bing is getting DALL-E 3 access even before OpenAI’s own ChatGPT does — that’s scheduled to happen this month, but only for paying users. Microsoft is likely to be the most popular image generating tool for a while.

“Microsoft is planning to use DALL-E tech in more than just Bing, too. It’s working on an AI image creation tool in the Paint app called Paint Cocreator, for instance, which will bring the DALL-E model right into Windows.”

Continue reading “You can now use the DALL-E 3 AI image generator inside Bing Chat” »

Northwestern University.

Like Flubber, the robot can’t bolt and strike against walls at lightning speeds – but it can walk. The mechanism of the robot, written by the AI system, works when air is pumped into its body. The rubbery robot’s legs contract, and continually pumping air into the robot causes it to slowly but steadily locomote.

Now that computer-generated imaging is accessible to anyone with a weird idea and an internet connection, the creation of “AI art” is raising questions—and lawsuits. The key questions seem to be 1) how does it actually work, 2) what work can it replace, and 3) how can the labor of artists be respected through this change?

The lawsuits over AI turn, in large part, on copyright. These copyright issues are so complex that we’ve devoted a whole, separate post to them. Here, we focus on thornier non-legal issues.

How Do AI Art Generators Work?

OpenAI, the leading artificial intelligence research lab, had an eventful week with major announcements that give insight into its future plans and direction. This was covered in the recent All-In Podcast:

Humane, a stealthy software and hardware company, is clearly milking the media hype cycle for all it’s worth. The company’s origin dates all the way back to 2017, when it was founded by former Apple employees Bethany Bongiorno and Imran Chaudhri. In the intervening half-decade, the firm has been largely shrouded in mystery, as it has put together the pieces of a mystery wearable, which it promises will leverage AI in unique ways.

The company’s been buzzy since it first engaged with the media — well before it offered the slightest bit of insight into what it’s been working on. In spite — or perhaps because — of such mysteries, Humane is now an extremely well-funded early-stage startup.

At the tail end of 2020, it raised a $30 million Series A at a $150 million valuation. The $100 million B round arrived the following September, including Tiger Global Management, SoftBank Group, BOND, Forerunner Ventures and Qualcomm Ventures. It all seemed like a strong vote of confidence for the still stealthy firm. This March, it went ahead and raised another $100 million.

“We hope that the research can contribute to and complement the arsenal of techniques used to diagnose breast cancer and to generate a large amount of data associated with it that may be useful in trying to identify large-scale trends that could help diagnose breast cancer early,” George added.

The team next plans to combine CBE techniques learned from professionals with AI and fully equip IRIS with sensors to determine the effectiveness of the whole system in identifying potential cancer risks. The ultimate goal is to have the manipulator detect lumps more accurately and deeper than it is possible only by applying human touch.

This promising development could revolutionize how women monitor their breast health. With safe electronic CBEs located in easily accessible places like pharmacies and health centers, women could have access to accurate results and take a proactive approach to their health.

Today’s generative AI revolution poses similar challenges and opportunities as the Industrial Revolution.

When Karl Wenner looks at his farm on Upper Klamath Lake in the mountains of southern Oregon, he sees a landscape in transition.

He and his partners converted part of their fields of barley into wetlands along the shore of the lake to filter runoff and protect the quality of the water that eventually flows back into the Klamath River, which empties into the Pacific on California’s coast. The project is part of a larger effort to clean up the river, remove dams and bring back salmon.

At Lakeside Farms, that transformation is being guided by a surprising source of information: the pollen collected by tens of thousands of honeybees. A Belgian start-up called BeeOdiversity enlisted Wenner, who is also a beekeeper, to help in a survey in the Klamath River Basin. Each colony, with 50,000 bees, harvests pollen over an area of more than two square miles, collecting as many as 4 billion tiny samples in a year. The resulting data creates a clear, accurate picture of the plant life and pollution present in the environment.