开启时会模糊预览图，关闭后正常显示
RT @voidzerodev: Ever since we announced Vite+, the excitement, and the questions, have been rolling in. We took some time to answer the mo…
Husband / Father of two / Founder @voidzerodev / Creator @vuejs & @vite_js. Chinese-only alt: @yuxiyou
BTW, for 2. it goes both ways, if it makes you look nice, you probably are.
Research Scientist @meta (FAIR), Prof. @Unige_en, co-founder @nc_shape. I like reality.
Tankies seethe when their crap aesthetics are called out I don't care about internal PRC legitimacy, the CPC can take care of that well enough without my counsel. The fact of the matter is simply that Stripping the Mandate goes hard
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
Order of my biggest AI capability updates based on new models - very circumstantial based on what I was thinking about / paying attention to: 1) GPT-3 - Claude 1 2) Claude 3.5 - o1 3) GPT-3.5 - GPT-4 4) nothing - GPT-2* 5) GPT-2 - GPT-3 6) AlphaGo and AlphaZero 7) o1 - o3 8) GPT-5 - Gemini 3 I think I missed the train on 4) which I think should have been #2 at least. I also probably would have updated harder on 5) if I had been doing what @gwern was doing at the time (lots of base-model prompting). I also think it was largely to Gwern’s credit that I saw some of the importance of this jump. Some credit to AI dungeon as well, which was a wild early look at capabilities for those who weren’t using OpenAI playground directly (if playground was even around then? I think it was but don’t remember) 1) was my ChatGPT moment because I was at Anthropic before ChatGPT was released publicly
Applying the security mindset to everything @PalisadeAI
Order of my biggest AI capability updates based on new models - very circumstantial based on what I was thinking about / paying attention to: 1) GPT-3 - Claude 1 2) Claude 3.5 - o1 3) GPT-3.5 - GPT-4 4) nothing - GPT-2* 5) GPT-2 - GPT-3 6) AlphaGo and AlphaZero 7) o1 - o3 8) GPT-5 - Gemini 3 I think I missed the train on 4) which I think should have been #2 at least. I also probably would have updated harder on 5) if I had been doing what @gwern was doing at the time (lots of base-model prompting). I also think it was largely to Gwern’s credit that I saw some of the importance of this jump. Some credit to AI dungeon as well, which was a wild early look at capabilities for those who weren’t using OpenAI playground directly (if playground was even around then? I think it was but don’t remember) 1) was my ChatGPT moment because I was at Anthropic before ChatGPT was released publicly
Applying the security mindset to everything @PalisadeAI
I have been enjoying LOS ANGELES this year. If you play tennis in the area, let's hit. I'll bring this guy, too.
https://t.co/zSf5Z2H78P https://t.co/ryMAyS77qn https://t.co/Gm6gdHaLgp On a mission to inspire 1B people to build stuff!
