Discover more from Machine Learnings
Projecting AI's role in the 2024 US Election
Along with real advances for musicians and medical professionals
We’re back with another edition of Machine Learnings, brought to you by the folks at Heyday.
Heyday is an AI-powered memory assistant that resurfaces content you forgot about while you browse the web.
Dream summer continues.
We’re still geeking out over LK-99. The results of last week’s breakthrough findings are not yet verified, but it hasn’t stopped partial replications and a host of excitement around what’s possible. Like, what would it mean to have a superconductor in your pocket?
What we're reading.
1/ The coolest AI drop of the week goes to Meta, who have packaged a generative audio suite with text-to-audio, a decoder, and code + model weights for your immediate use. This is a high quality tool. Learn more at the Meta AI Blog >
2/ The media loves talking to Sam Altman, and he’s been talking about the downsides of next year’s US presidential election. We have problems to solve. Learn more at Quartz >
3/ LLMs have the ability to rapidly generate
believable text, so when applied to something like travel, we can imagine a library of guides popping up quickly. Unfortunately, scammers have taken hold. Buy one, and you may end up in the wrong place. Learn more at The New York Times >
4/ [Listen] Holden Karnofsky is the cofounder of Open Philanthropy and the partner of Anthropic’s President, so his views are both informed and unique. Here he dives into AI risks and the playbook we can run to keep AI’s impact on the safer side. Learn more at 80000 Hours >
5/ Drug design goes well beyond our earliest advances in protein folding. Here’s more on one company’s efforts to design a solution for schizophrenia. Learn more at Fast Company >
6/ More medical advances. This time from Google’s work synthesizing recent research and applying them to a foundational model for medicine. Learn more at Google Research >
7/ Formal US AI policy is coming, and we’re still wondering what that will look like. Here’s a great deep dive looking back to the US government’s privacy legislation and how that may impact what we’ll end up with. Learn more at WIRED >
Research for this edition of Machine Learnings was enhanced by Heyday, the AI-powered memory assistant.
Thanks for reading Machine Learnings! Subscribe for free to receive new posts and support my work.