OpenAI Meltdown: Artificial Intelligence is no match for Human Stupidity

Tech is clean, humans are messy. The Board of OpenAI, charged with the grave responsibility of overseeing development of one of the most important technologies in human history, on Friday decided oh, f*** it, let’s play Game of Thrones! The three outside directors of the 6-person Board, all very concerned about AI safety, managed to flip co-founder and Chief Scientist Ilya Sutskever to the view that CEO Sam Altman was commercializing too fast, so they voted Sam off the island, and demoted Board Chair and co-founder Greg Brockman to Third Assistant Batboy (or equivalent.) Surprising no one, except the rash and clueless rogue Rump Board, Greg quit.

Chaos ensued. The Rump Board proved, over and over again, that it was reckless, feckless, and clueless. First, they had blindsided Microsoft, who had invested $13 billion in the company. Satya Nadella, CEO of Microsoft, gave a master class in grace under pressure through the weekend, mediating the dispute between Sam and the Rump Board. Then the Rump Board hired a new CEO, Emmett Shear, a gamer billionaire with the facial hair of a supervillain. (Their top 3 picks - including the CEO of competitor Anthropic - all turned them down flat. Shear was Choice #4.) (Fun fact: Emmett and Sam were both in the initial class of founders at legendary startup incubator Y Combinator.)

At that point, Sam and Greg were set to move to Microsoft and set up a rival AI research lab. Then OpenAI’s employees revolted. At last count, 745 of the 770 OpenAI employees have signed onto a letter threatening to resign unless Sam was reinstated and the Rump Board resigned. Do the OpenAI employees love Sam? Well, yes. He’s led them on the most exciting journey of their professional lives, taking the company to Olympian heights. More relevant, perhaps, Sam had arranged a sale of employee shares that looked to make many of them instant millionaires. And the Rump Board shenanigans had now made OpenAI radioactive to investors, blowing up the planned sale. Even employees who were lukewarm on Sam hated having their stock made worthless, so the Rumpistas had to go.

At this point, the Rump Board threw in the towel. Sam and Greg are set to come back. Two functioning adults will join the Board, Former Salesforce CEO Bret Taylor as Chair, and former Treasury Secretary and President of Harvard Larry Summers as, one supposes, Eminence Grise. Ilya will step down from the Board, but keep his job as Chief Scientist (for now.) Two other Rumpistas will resign from the Board, but one will remain, Adam D’Angelo, CEO of Quora, and by all accounts, a stubborn mule of a man. So, the drama may not be over.

Tweet from Greg on how the firings went down

The Return of the King: How Sam Altman fired the Board that fired him

Loser: Open AI

Before Friday, OpenAI was the undisputed king of the AI hill, inventing tech and shipping product at a pace no one else could match. Now, it looks like a house of cards, ready to fly apart at any moment. Perception matters, and the company’s very public Board food fight will scare away customers and investors, and embolden competitors.

Winner: Anthropic (and every other OpenAI competitor)

Anthropic, developer of ChatGPT’s chief competitor, Claude, was founded by breakaway researchers from OpenAI who thought that the company wasn’t paying enough attention to AI safety, the issue that sparked this recent OpenAI Board revolt. Anthropic found that it, too, had to get big investors to fund AI research and operations, which meant that it too had to commercialize for revenue, so at this point any difference in AI safety seems mostly a talking point and an ad slogan. Seizing the moment, Anthropic launched a significant upgrade, Claude 2.1, announcing it as being “developed by a company that isn’t actively at war with itself.” Zing!

Anthropic releases Claude 2.1 with a bit of sass

Best Comeback: OpenAI’s brilliant and undaunted staff

Amid the Board meltdown, when all of OpenAI’s 778 employees’ jobs, careers, and millions in stock options were on the line, these badasses decide that they would just, you know, launch another feature for the ChatGPT phone app. Free users will get the Voice feature, which allows you to talk to the app and get voice responses. The demo even has a veiled joke about the Board drama, Ho hum. Just another Tuesday…

ChatGPT Voice announcement and demo

Forward Health launches AI-powered, doctor-free “CarePods”

Ex-Googler Adrian Aoun wants to scale healthcare. His first try as CEO of healthtech startup Forward Health was a chain of tech-forward doctor’s offices which had an Apple Store vibe. Like many a healthtech entrepreneur before him, he found that person-to-person medical service just doesn’t scale. So how to get the people out? Say hello to the new doctorless “Carepod”, a self-contained, standalone medical station reminiscent of airport lactation pods for nursing mothers. The whole thing is run by a medical AI that asks questions and provides various types of medical tests. If a prescription is needed, there are physicians online that can provide them. Access to the pods is by subscription, currently $99/month. Although the current offerings of the Carepod are limited, CEO Aoun has big ambitions to be able to replace almost all of current healthcare over the next 20 years or so, up to and including heart surgery.

AI-powered Carepods for your health care

Mayo Clinic develops AI to detect pancreatic cancer

Pancreatic cancer is notoriously deadly, because it is usually found too late. Mayo Clinic has developed an advanced AI that can detect pancreatic cancer better than the best radiologists, using only standard CT scans.

Mayo Clinic develops an AI to detect early pancreatic cancer