2024-09-05T18:31:30+00:00 | 🔗
RT @dylfreed: The new Qwen2-VL-7B Instruct model gets *100%* accuracy extracting text from this handwritten document. This is the first ope…
2024-09-05T15:40:43+00:00 | 🔗
Omg I thought I was crazy for thinking that people were scaling crazy things… Glad Ilya agrees https://t.co/IQ64tPsBaO
2024-09-05T04:07:11+00:00 | 🔗
AI is literally making life worse rn because of BAD customer service automation. I hate talking to the chatbots. Horrible. Way worse than DOOM.
2024-09-04T13:29:48+00:00 | 🔗
AI deep state is the real outcome… not DOOM https://t.co/H9XOje6yqF
2024-09-03T02:28:06+00:00 | 🔗
A good test of AI-generated applications (artifacts, v0 things, replit modal things) will be whether or not the AI can deploy it to a custom domain. I'm not talking about like automating purchasing on and using route53, or even web agents on NameCheap. I'm talking about IT.....
2024-09-01T20:24:01+00:00 | 🔗
Yes but also it’s just that Math and Code live in a virtual sandboxes world Most everything else does not https://t.co/fiUMCXs34m
2024-09-01T17:33:23+00:00 | 🔗
It's never, "How to add a feature to your existing application". Thonk https://t.co/SxGx5MTmWs
2024-08-30T19:37:09+00:00 | 🔗
Or that AI tools are hard to use in an existing code base, or worse a whole crazy infra stack… https://t.co/MxGyKevUjQ
2024-08-30T17:37:44+00:00 | 🔗
RT @drGregBowman: @Brcremer Seriously. These people could at least have some internal consistency and apply Biblical principles for choosin…
2024-08-30T14:57:06+00:00 | 🔗
How can humans and AI compete on ENERGY??? The human brain is insanely more efficient than any AI. Bang for your buck on an energy basis is wildly in favor of humans, hello?