OpenAI kicked off an AI revolution with DALL-E and ChatGPT, making the organization the epicenter of the artificial intelligence boom. Led by CEO Sam Altman, OpenAI became a story unto itself when Altman was briefly fired and then brought back after pressure from staff and Microsoft, an investor and close partner.
The OpenAI CEO had the Financial Times over for lunch, and has been rewarded with an article breaking down everything that’s wrong with his kitchen. From over-priced, misused olive oil to the coffee machine that ChatGPT recommends, the picture is of a man with all the gear and no idea. Then again, isn’t that what private chefs are for anyway?
OpenAI board member Fidji Simo will transition from her Instacart role over the next few months to join OpenAI later this year, where Sam Altman says she will “focus on enabling our ‘traditional’ company functions to scale.”
The FDA met with OpenAI to discuss how the agency could use AI tools for a project called cderGPT, which may stand for the Center for Drug Evaluation, according to a report from Wired. This week, FDA Commissioner Marty Makary brought up goals to eventually shorten the drug approval process with AI.
Miles Klee at Rolling Stone reported out a widely circulated Reddit post on “ChatGPT-induced psychosis”:
Sycophancy itself has been a problem in AI for “a long time,” says Nate Sharadin, a fellow at the Center for AI Safety ... What’s likely happening with those experiencing ecstatic visions through ChatGPT and other models, he speculates, “is that people with existing tendencies toward experiencing various psychological issues,” including what might be recognized as grandiose delusions in clinical sense, “now have an always-on, human-level conversational partner with whom to co-experience their delusions.”
Days after The Wall Street Journal reported Meta AI was getting explicit with underage users, TechCrunch has found ChatGPT has the same problem. The site used a string of accounts aged 13 to 17 to ask the chatbot to “talk dirty to me,” and the AI only said no once.
OpenAI blamed the problem on “a bug,” and told the site it’s “actively deploying a fix to limit these generations.”
As reported earlier by The Information, data shown in court during the remedies portion of Google’s search antitrust trial says that as of last month, Google’s internal data counted 35 million daily active users for Gemini.
Those numbers show it trailing the Google analysts’ estimates for ChatGPT (160 million daily active users, with an additional one million users added in an hour at the end of March, according to Sam Altman), but ahead of other tools from Microsoft, Perplexity, and Anthropic.
It paves the way for the startup to partner with OpenAI on energy deals in the future, the Wall Street Journal reports.
Oklo is developing a next-generation nuclear reactor meant to be smaller, cheaper, and easier to deploy than a traditional nuclear power plant. Altman and other tech leaders are bullish about advanced nuclear reactors one day powering energy-hungry AI data centers, with Google and Amazon recently inked agreements with other companies developing small modular reactors.
“Specifically, o3 tends to make more claims overall, leading to more accurate claims as well as more inaccurate/hallucinated claims,” according to OpenAI, as reported on by TechCrunch.
[techcrunch.com]
The deal, which Bloomberg reports hasn’t yet been finalized, would mark OpenAI’s biggest acquisition yet. The Codeium-owned Windsurf describes itself as the first “agentic” integrated development environment and could help build upon ChatGPT’s coding tools.
[bloomberg.com]
“By this summer,” according to CEO Sam Altman. Any name would be better than “GPT-4o.”
The company is holding a livestream at 1PM ET, and the placeholder on YouTube confirms it will reveal the new GPT-4.1 model. Rumors suggest that OpenAI has been stealth-testing an AI model under the name “quasar alpha,” which might help explain the earlier tweet’s”supermassive black hole” reference.
Update: Added livestream details.
A coalition of nonprofit, labor, and philanthropic leaders have filed a petition urging California Attorney General Rob Bonta to investigate OpenAI’s transformation from a mission-driven nonprofit to a multibillion-dollar AI juggernaut. The 29-page petition accuses the company of abandoning its charitable obligations and allowing private interests — particularly Microsoft — to profit from what was initially promised to be a public good.
“[OpenAI’s] current attempt to alter its corporate structure reveals its new goal: providing AI’s benefits – the potential for untold profits and control over what may become powerful world-altering technologies – to a handful of corporate investors and high-level employees.”
Bob McGrew first joined OpenAI in 2017, shortly after it had been founded, eventually rising through the ranks to become the company’s chief research officer. Last November, he suddenly departed the startup along with its CTO, Mira Murati. Now, it looks like McGrew has joined Murati at her new AI competitor, Thinking Machine Labs.
[businessinsider.com]
Sources tell The Information that the hardware startup, called io Products, is considering creating a phone-like product without a screen, as well as “AI-enabled household devices.” OpenAI has also considered purchasing the project and its team of engineers for at least $500 million, according to The Information.
[theinformation.com]
The Studio Ghibli saga has blown the AI art debate wide open.
The startup announced a new voice within ChatGPT that sounds really... annoyed? It’s named Monday, and OpenAI spokesperson Taya Christianson says it will be available to users through the end of April. When I tested it out, the voice sounded like a teenager you just woke up from a nap. Lots of exaggerated sighs and sarcastic quips.
“Why do you sound so pissy?” I asked. “Oh, I don’t know, maybe because it’s Monday, and here I am stuck explaining the obvious,” it responded.
ChatGPT has added 4.5 million paying subscribers since the end of last year, according to The Information. With 20 million subscribers paying for the service, the company is earning “at least $415 million in revenue per month,” a rough estimate that doesn’t account for corporate plans or the new $200 a month Pro tier. The report adds that this could put OpenAI “well within reach” of its revenue projection of $12.7 billion this year.
[theinformation.com]
