Weekly Top Picks #40
Or the new "What You May Have Missed"
Hi guys!
I'm writing from Cantabria, north Spain, where the plains are of deep, vivid green, the beaches are long and beautiful, the sea is cold yet always welcoming with its gentle waves, and the sun plays hide and seek with the omnipresent clouds, ranging from heavenly white to threatening dark gray.
I haven’t published anything since Monday last week. I hope the above description explains why! I have the (longish) second and third parts of the AI hurdles series ready for Wednesday and Friday. Today, I bring you a light version of “What You May Have Missed” as a curated list of the most important events of the week, without categories.
I’m considering this Top-N Picks format to replace the current WYMHM weekly review. I feel that as it stands now, it gets overwhelming (that’s the last effect I want it to have). Maybe a double- or triple-curated list that highlights the most important events is better (i.e., I find ~100 links, curate it down to ~25, then curate it further to ~5-10).
On the other hand, it may also be that by the end of the week, you have already read about this elsewhere. I will think about how to solve this trade-off to maximize both the value I provide and conciseness—one possibility is to add richer context or brief commentary, which daily newsletters so often lack.
Anyway, before you click on a normal subscription link, let me remind you that the 30% off Summer Special Offer is still on (expires in three weeks):
Without further ado, this week’s top picks at a glance:
ChatGPT leans liberal—but an independent analysis has found the methodology questionable.
Newly-founded Sakana AI defies the notion that generative AI (and LLMs like ChatGPT) is our best chance at creating intelligent systems.
The copyright wars: How the NYT could indirectly force OpenAI to delete ChatGPT and start over.
How Google DeepMind plans to beat OpenAI and ChatGPT with Gemini—Google co-founder Sergey Brin is involved.
The pause letter wasn't fully supported by many signatories—Not all “AI doomers” are actually doomers after all.
Generative AI is officially in the “Peak of Inflated Expectations,” says consulting firm Gartner. When will it reach the “Plateau of Productivity?”
Art created by AI is not copyrightable, according to a judge, but the ruling applies exclusively to AI authorship.
New paper: “There are no obvious barriers to building conscious AI systems.” But should we do it?