Discover more from Machine Learnings
Curation and shorthand
+ the latest announcements from Google I/O
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.
Thanks for reading Machine Learnings! Subscribe for free to receive new posts and support my work.
Occasionally, we read things that take hold over our brains and ask us to rethink the future. This week, it was Venkatesh Rao’s speculation around what happens to language after infinite tailoring.
Might this look like shorthand? A language that each individual has innate understanding of, but is less recognizable to others? If so, what can we pull from to design software that helps us tap into our shorthand?
These are some of the questions we’re asking as we build tools for your memory. We don’t just want you to remember; we want you to apply that unique insight right when it’s needed.
We think these patterns are critical for creating tools that work for us, individually. If you’re interested in what that looks like in your life, reach out or try it free for the next two weeks.
What we're reading.
1/ Our longest read of the week covers AI and…acupuncture? I was as curious as you are, and this did not disappoint. Open up the black box and consider the crossovers in designing experiments for understanding AI (and acupuncture). Learn more at WIRED >
2/ Couple with this – OpenAI’s attempt to understand what the individual components (what we call ‘neurons’ and ‘attention heads’) are doing within LLMs. Learn more at OpenAI >
3/ We’re big hip hop fans at Heyday HQ, so comparing the peak writing process to the requisite understanding of producing timeless beats hit close to home. This is curation. Well worth the read from Stephen Marche. Learn more at The Atlantic >
4/ Google I/O was this week, and we’ll spotlight a few of the releases. Most of the AI announcements are embedding generative AI across Google’s footprint, but we can also see advances in their new model, PaLM 2. Learn more at Engadget >
5/ Is this the end of StackOverflow? The numbers don’t look friendly… Learn more at Similar Web >
6/ A follow-on profile on Geoffrey Hinton. This is an interesting case for us, because while we don’t sit on the doom side, widening our perspectives have proven to help us see a clearer picture of what’s going on. Learn more at WIRED >
7/ Author and power software user, Steven Johnson, has been working with Google to progress the tools for thought space, and their announcement of Project Tailwind brings semantic search to your personal library. We’re intrigued… Learn more at Steven Johnson’s Substack below.
Research for this edition of Machine Learnings was enhanced by Heyday, the AI-powered memory assistant.