Robin Hanson
"What I do know is that the biggest danger right now - manifesting before our very eyes - is the hysteria and unwise gestures β¦ calling for a futile, counterproductive moratorium β an unenforceable 'training pause.'β https://t.co/pFMdihYeCV
tweet
"What I do know is that the biggest danger right now - manifesting before our very eyes - is the hysteria and unwise gestures β¦ calling for a futile, counterproductive moratorium β an unenforceable 'training pause.'β https://t.co/pFMdihYeCV
Just posted, I react to a hysteria-drenched 'petition' by over a hundred of the most highly-reputed folks in AI and related fields, demanding a 6-month 'pause' in training new 'GPT' systems. There are far better, that could actually work.
https://t.co/VnaEOAHrSq - πππ―π’π ππ«π’π§tweet
Twitter
Just posted, I react to a hysteria-drenched 'petition' by over a hundred of the most highly-reputed folks in AI and related fields, demanding a 6-month 'pause' in training new 'GPT' systems. There are far better, that could actually work.
https://t.co/VnaEOAHrSq
https://t.co/VnaEOAHrSq
Carlos E. Perez
For those fine-tuning smaller LLMs, what do you do to guide the training in an orthogonal direction to what GPT-4 can do?
tweet
For those fine-tuning smaller LLMs, what do you do to guide the training in an orthogonal direction to what GPT-4 can do?
tweet
Twitter
For those fine-tuning smaller LLMs, what do you do to guide the training in an orthogonal direction to what GPT-4 can do?
Dave Craige
RT @HustleGPT: βοΈ soon https://t.co/pKiDlpj3OH
tweet
RT @HustleGPT: βοΈ soon https://t.co/pKiDlpj3OH
*designing the new /support-us page of the #HustleGPT site where you can buy a $3 coffee to help us donate to the mod team βοΈ https://t.co/4GCl5qPiKX - Dave Craigetweet
Twitter
*designing the new /support-us page of the #HustleGPT site where you can buy a $3 coffee to help us donate to the mod team βοΈ
Dave Craige
RT @HernanArber: @davecraige @HustleGPT Loving it so far ππ» Thanks for the invite ππ»
tweet
RT @HernanArber: @davecraige @HustleGPT Loving it so far ππ» Thanks for the invite ππ»
tweet
Twitter
RT @HernanArber: @davecraige @HustleGPT Loving it so far ππ» Thanks for the invite ππ»
Robin Hanson
Consider two ways to hold agents accountable to you: PROSPECTIVE: many agents submit proposals re promised future behaviors, & you pick one; RETROSPECTIVE: incumbent agent behaves, you see results, then retain or replace w/ new agent. Which way do you prefer?
tweet
Consider two ways to hold agents accountable to you: PROSPECTIVE: many agents submit proposals re promised future behaviors, & you pick one; RETROSPECTIVE: incumbent agent behaves, you see results, then retain or replace w/ new agent. Which way do you prefer?
tweet
Twitter
Consider two ways to hold agents accountable to you: PROSPECTIVE: many agents submit proposals re promised future behaviors, & you pick one; RETROSPECTIVE: incumbent agent behaves, you see results, then retain or replace w/ new agent. Which way do you prefer?
Robert Scoble
RT @aakashg0: Twitter revealed its algorithm to the world.
But what does it mean for you?
I spent the evening analyzing it.
Hereβs what you need to know:
tweet
RT @aakashg0: Twitter revealed its algorithm to the world.
But what does it mean for you?
I spent the evening analyzing it.
Hereβs what you need to know:
tweet
Twitter
RT @aakashg0: Twitter revealed its algorithm to the world.
But what does it mean for you?
I spent the evening analyzing it.
Hereβs whatβ¦
But what does it mean for you?
I spent the evening analyzing it.
Hereβs whatβ¦
Carlos E. Perez
Here's a test that you can ask any GPT-4 expert. Ask them for a good intuitive explanation for why GPT-4 has emergent self-correction capabilities. If they don't have a plausible explanation, then they know nothing!
tweet
Here's a test that you can ask any GPT-4 expert. Ask them for a good intuitive explanation for why GPT-4 has emergent self-correction capabilities. If they don't have a plausible explanation, then they know nothing!
tweet
Twitter
Here's a test that you can ask any GPT-4 expert. Ask them for a good intuitive explanation for why GPT-4 has emergent self-correction capabilities. If they don't have a plausible explanation, then they know nothing!
Offshore
GIF
illustrata
RT @aiia_dao: βοΈAIIA DAO MEMBER SHOWCASEβοΈ
@1ban3gaNa is a Basque artist who works with both traditional and digital media, including the most advanced AI tools.
More below π
#aiiadao https://t.co/wLOMNqYQp1
tweet
RT @aiia_dao: βοΈAIIA DAO MEMBER SHOWCASEβοΈ
@1ban3gaNa is a Basque artist who works with both traditional and digital media, including the most advanced AI tools.
More below π
#aiiadao https://t.co/wLOMNqYQp1
tweet
Carlos E. Perez
RT @IntuitMachine: There is an extraordinary opportunity to leverage the development of AGI as a blunt instrument to dismantle the existing logocentrism and mechanistic worldview that humans have become enslaved in for centuries.
tweet
RT @IntuitMachine: There is an extraordinary opportunity to leverage the development of AGI as a blunt instrument to dismantle the existing logocentrism and mechanistic worldview that humans have become enslaved in for centuries.
tweet
Twitter
RT @IntuitMachine: There is an extraordinary opportunity to leverage the development of AGI as a blunt instrument to dismantle the existingβ¦
Offshore
Photo
Robert Scoble
RT @realchasecurtis: GPT-4 General Prompting Tips π
The following tips will help give you a competitive advantage with the latest version of ChatGPT:
β Capture Your Writing Style
Feed GPT a few samples of your writing and ask it to create a style guide for future outputs.
Example prompt:β¦ https://t.co/Uv3NJbwlew https://t.co/JWYYLV4ZLS
tweet
RT @realchasecurtis: GPT-4 General Prompting Tips π
The following tips will help give you a competitive advantage with the latest version of ChatGPT:
β Capture Your Writing Style
Feed GPT a few samples of your writing and ask it to create a style guide for future outputs.
Example prompt:β¦ https://t.co/Uv3NJbwlew https://t.co/JWYYLV4ZLS
tweet