🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

1/ How can I put this? "Expected utility maximizers don't maximize utility." Why? Because utility is not usually an ergodic quantity in the mathematical models used by economists, and maximizing its expected value doesn't mean much in the real world.

wild I hadn’t read this earlier! So rare nowadays for a manuscript to exhibit such snark and bile toward its named intellectual enemies — extremely refreshing

what if instead of an intelligence explosion it seems like we're getting an intelligence implosion

This fascinating new research overturns longstanding assumptions. It finds that lower-income groups are more concerned about the environment, and prefer environmental protection over economic growth, compared to higher-income groups. Data from the USA.

Creativity is the ability to make intelligent decisions without a destination in mind. That's why training LLMs to solve problems (the pre-specified destination of the chain of thought) will not lead to creativity.

For those who like music, there was once an interesting album released in November 1999 by a band called Rage Against The Machine. An album full of interesting lyrics. Well worth a listen!

New paper: "Large Language Models & Emergence: A Complex Systems Perspective" (D. Krakauer, J. Krakauer, M. Mitchell). We look at claims of "emergent capabilities" & "emergent intelligence" in LLMs from perspective of what emergence means in complexity science. ⬇️

It’s a hefty 206-page research paper, and the findings are concerning. "LLM users consistently underperformed at neural, linguistic, and behavioral levels" This study finds LLM dependence weakens the writer’s own neural and linguistic fingerprints. 🤔🤔 Relying only on EEG,

in other news, using an escalator instead of walking up stairs significantly decreased cardiovascular function and leg strength.

Obviously using chatbots makes you more dumb, everyone is skipping over the funniest part: the study authors are such dedicated LLM haters they set up traps to expose AI users lol

Helen De Cruz (1978-2025) | Such sad news. She was a wonderful and amazing person. dailynous.com/2025/06/20/hel…

Helen De Cruz left our world today. Her book "Wonderstruck: How Wonder and Awe Shape the Way We Think" reminded me why I love doing science so much. Helen had agreed to come on the Night Science Podcast but sadly was not feeling well enough in the end for the recording.

RIP Helen De Cruz (1978-2025) 💔 Helen De Cruz — Professor of philosophy at Saint Louis University, has died. dailynous.com/2025/06/20/hel…

I am devastated to learn of my dear friend and fellow philosopher Helen De Cruz’s passing. With Helen's blessing, her husband Johan and I agreed to circulate a GoFundMe on their children’s behalf. Please feel free to share and circulate widely. gofundme.com/f/support-hele…

I wrote a short piece remembering Helen De Cruz, who passed away yesterday. philipgoff.substack.com/p/helen-de-cru…

Helen de Cruz (1978-2025), RIP open.substack.com/pub/digression…@

I knew Helen De Cruz only in passing but she was so clearly a person of substance, compassion, and character

Reward models (RMs) are the moral compass of LLMs – but no one has x-rayed them at scale. We just ran the first exhaustive analysis of 10 leading RMs, and the results were...eye-opening. Wild disagreement, base-model imprint, identity-term bias, mere-exposure quirks & more: 🧵

Anthropic or Anthropic-sponsored safety papers

Grok going full Nazi because its owner didn’t like it answering honestly is the future of AI. Have fun, kids.