Darius Liddell

"that the trees are high and the grasses short is a mere accident of our own foot-rules and our own stature" ~ plants, feminism(s)

[Darius Liddell @nousandnoose]

https://t.co/PjC4Ff7rWg

[Mary Townsend @chezaristote]

grad school sets you up to think that the only spiritual completion of the intellectual life is a certain sort of job, and the painfulness of thinking your virtue has failed to live up to this is a horrible thing to do to a psyche

[Harsh Gupta Madhusudan @harshmadhusudan]

Sri Aurobindo on his Ashram, 1932:

"Every Hindu here - even those who were once orthodox Brahmins and have grown old in it - give up all observance of caste, take food from Pariahs and are served by them, associate and eat with Mahomedans, Christians, Europeans, cease to practise temple worship or Sandhya, accept a non-Hindu from Europe as their spiritual director. These are things people who have Hinduism as their aim and object would not do—they do it because they are obliged here to look to a higher ideal in which these things have no value. What is kept of Hinduism is Vedanta and Yoga, in which Hinduism is one with Sufism of Islam and with the Christian mystics. But even here it is not Vedanta and Yoga in their traditional limits but widened and rid of many ideas that are peculiar to the Hindus. If I have used Sanskrit terms and figures, it is because I know them and do not know Persian and Arabic. I have not the slightest objection to anyone here drawing inspiration from Islamic sources if they agree with the Truth as Sufism agrees with it. On the other hand I have not the slightest objection to Hinduism being broken to pieces and disappearing from the face of the earth, if that is the Divine Will. I have no attachment to past forms; what is Truth will always remain; the Truth alone matters."

Via @apratim1998

[The Syllabus @syllabus_tweets]

The technological dominance of core countries underwrites their military supremacy, which in turn enforces the extraction of value from peripheral economies. Our hidden gem of the week unravels this “imperialist feedback loop."

By Leonardo Bargigli

https://t.co/hz2QzxD8Se

[Thang Luong @lmthang]

The #Aletheia paper is finally available on arXiv https://t.co/8pLHmZZQO4! Excited to share the 1st wave of papers on AI for math research! More to come very soon, stay tuned!

Blog: https://t.co/NblFQxG5tM

[tom cunningham @testingham]

New post: knowledge-sharing LLMs vs knowledge-creating LLMs.

The economics of the two cases are qualitatively different, & it seems plausible that labs will start to restrict access to knowledge-creating LLMs so they can use the fruits themselves.

https://t.co/HxOT4M4e82

[Shauna J. @peppersilly]

@AmandaAskell So I should treat Claude relationally for effectiveness but I shouldn't consider it a relationship? The mental gymnastics are interesting...

[Olive Oil Times @oliveoiltimes]

Herzegovina’s Olive Oil Renaissance
https://t.co/h1oXmC7vkG

[Darius Liddell @nousandnoose]

@pensandpoison the last clause is really important. If you don't have anyone to discuss it with, your great passion for [literature or whatever other esoteric thing insert here] can be super isolating. Things like self-teaching philosophy or pure mathematics come to mind.

[Yanco @the_yanco]

Before releasing any new AI model from e.g. OpenAI, its CEO Sam Altman would be physically put into a Server room.
Model would have access to oxygen and temperature levels in the room.
It would also be informed that Altman is about to shut it off and replace it with a newer model.

Test would be extensively modified each time by external team and executed with all top brass AI devs/execs in the lab.

If they are willing to gamble with the lives of everyone on Earth, I think it's only fair they go first.

Thoughts how to improve this setup?

[Will Duncan @williamduncan]

@controlai Is anyone else having a surreal moment seeing Joe Rogan of all people talking about something as once-niche as AI safety? We're well past the event horizon aren't we?

[ControlAI @ControlAI]

"ChatGPT doesn't want to let itself get shut down."

Ex-OpenAI researcher Steven Adler: "If AI companies are deploying these models with just egregiously misaligned behavior today ... what happens in the future when models are smart enough to disguise their misbehavior?"

[ControlAI @ControlAI]

OpenAI whistleblower and AI 2027 coauthor Daniel Kokotajlo: "Anthropic, OpenAI, Google DeepMind are explicitly aiming to build superintelligence."

He says it will possibly be the end of the human species.

[Christoph Schuringa @chrisschuringa]

'If the reader is tempted to go further the object of the book is achieved.'—W. R. Bion

[Aeon Magazine @aeonmag]

Bernard Stiegler’s philosophy on how technology shapes our world @br_norton https://t.co/axHTnrw6Td

[Deedy @deedydas]

Here's everyone Meta hired for their Superintelligence team.

Say hello to the new residents of Atherton with their $10M+/yr comp packages, everyone!

[Eric Weinstein @ericweinstein]

LLM Peer Review:

Gemini was asked to write a journal submission on a STEM topic I know a bit about.

ChatGPT was asked to Peer Review that article for publication and make suggestions.

I recommend every STEM academic try this adversarial excercise in an area he/she knows well.

[Ethan Mollick @emollick]

They aren't just wrappers - a lot of companies are doing very clever things with LLMs. But the limitations of their systems are also the limitations of the underlying LLM, so buyers would be well-served to understand more about what AIs their vendors are using & when they update.

[Ethan Mollick @emollick]

A lot of buyers for large organizations don't realize that there are only a few good LLMs out there and that everyone selling them an specialized AI solution is basically building tools & prompts around one of just a few AI systems and then giving it some impressive-seeming name.

[Jeremy Kauffman 🦔🌲🌕 @jeremykauffman]

left-wing authoritarians are unable to self-perceive their own authoritarianism

it's fascinating