Hello, fellow humans! You’re reading our limited series of Saturday Daily Briefs. While it’s focused on AI, it’s curated, written, and edited by actual people.
Got some questions about AI you’d like answered? Or just some AI hallucinations you’d like to share? Email us anytime. Enjoy!
Here’s what you need to know
OpenAI added Microsoft to its board as a nonvoting member. The seat will give one of the generative AI company’s biggest investors a bird’s-eye view into its happenings.
US officials are concerned about one AI giant’s ties to China. G42, a quickly growing UAE firm that has etched partnerships with European drugmaker AstraZeneca and OpenAI, has been flagged by the CIA and other American intelligence agencies.
The US, UK, and a dozen other countries made an AI pact on safety. The nonbinding international agreement, though among the first of its kind, is pretty much toothless.
The largest medical association in the US published AI guidelines for doctors. It highlighted the lack of government policies (pdf) around the matter, adding that hospitals should share in some of that oversight.
AI used AI to make a robot. Right now it looks like a squishy purple blob, but it can move.
Amazon has a Q for you
Amazon, potentially feeling left out of the Great Chat Race, decided to roll out an AI chatbot for work this week by the name of Q. To be honest, Q kind of sounds like middle management: workers can talk with it and tell it to solve problems and do stuff. The idea is that Q will pretty much know the ins and outs of the job, as it’s powered by an employer’s own data.
Q is stepping into a crowded AI chatbot arena at a slightly lower price than competitors. It starts at $20 per month per user, so, a hefty cost for larger companies, but lower than Microsoft Copilot, which starts at $30 per month per user.
To be sure, many people are already using generative AI in the workplace (54% of Daily Brief readers have some thoughts about that below!), so what’s it to Q? Well, its feelings probably depend on whether it identifies more with James Bond or Star Trek.
Quotable: Wage against the machine
“If work does not vanish in the age of AI, then poorly paid jobs will most certainly proliferate, I argue.”—Peter Fleming, a professor of business and society at City University of London, in a 2019 paper (pdf) in which he proposed the idea of “bounded automation.”
Fleming’s prediction of worsening wages in the age of AI got some backing from the European Central Bank this week. It published research that found while AI hasn’t yet come for jobs, its effects on wages are leaning toward the negative.
This one (really bad) trick
Here’s an idea that definitely won’t boost your conference’s diversity numbers: listing a lineup of fake female panelists.
DevTernity, an annual online software conference, was canceled after observers noticed something amiss with its agenda. Multiple women speakers on its lineup didn’t seem to actually exist. The fallout was swift, with executives from the likes of Microsoft and Amazon canceling their speaking engagements. Eduards Sizovs, who organized the event, said a profile was auto-generated as a placeholder, but it wasn’t a good look for a field that has long struggled with diversity. (We wrote about this in our latest edition of the Memo, a free email about work that is also written by a human. Sign up here!)
Other great AI reads
👀 What Nvidia’s earnings say about the future of generative AI
🍎 Apple’s succession plan is completely different to OpenAI’s, in that it actually has one…
🤦 …though after the Sam Altman fiasco, OpenAI is making changes to its board.
🙀 The Pentagon’s AI initiatives are accelerating hard decisions on lethal autonomous weapons
🇷🇺 Putin wants to boost AI work in Russia to fight a Western monopoly he says is “unacceptable and dangerous”
Ask an AI to analyze what we asked you
Last week, we asked ChatGPT to celebrate its birthday by making a survey that would help us understand how our readers use generative AI in their daily lives.
ChatGPT helpfully offered to analyze it for us, so we fed it an .xlsx of the results from SurveyMonkey, which consisted of 351 responses to our 8-question survey. Then we asked some questions like “What are general trends?” and, “How often do our readers use AI?” and got some nicely written answers that might go well on, say, a company report.
It’s easy to see how analyzing a much larger, more complicated survey might benefit from this kind of treatment, and everything was going fine, until we asked ChatGPT what types of generative AI you used.
“It appears that there is no data available in the dataset for the question related to the types of generative AI used most frequently. This could be due to the way the data was organized or labeled in the survey,” it said.
Well, our first question was, “Which types of generative AI do you use? Select all that appl” so we’re calling that a hallucination. Your responses were, if you’re curious, 68% text generation, 27% image generation, 10% data synthesis, 3% video generation or editing, 3% music or sound generation, and then 22% of you do not use it at all.
The most common frequency of usage was several times a week (22%) and most of the people who use AI use it for work (54%).
And, perhaps the most important result of all, 56% of you were still on the fence about whether AI will become sentient, stop being so cheerful all the time, and turn our lives into the stuff depicted by the more nightmarish images it generates. ChatGPT said it just thought it was an interesting question!
Did you know we have two premium weekend emails too? One gives you analysis on the week’s news, and one provides the best reads from Quartz and elsewhere to get your week started right. You can get those by becoming a member—and take 20% off!
Our best wishes for a very human day. Send any news, comments, .xlsx datasets, and questions to ask ChatGPT about the datasets to [email protected]. Reader support makes Quartz available to all—become a member. Today’s AI in Focus Daily Brief was brought to you by Michelle Cheng, Morgan Haefner, and Susan Howson.