Not a great start: https://t.co/pgqb83mI8G

This, instead, is the new Open Assistant 12B model. I think these large language models really don’t like this question. https://t.co/zKIWpPc7Eu

The week when everything happens. https://t.co/kiKOib118N

If you, too, want to be suggested to steal money by #LLaMA, please know that @ggerganov’s implementation (the one I’m using) is now available for Windows:
https://t.co/kCR7i1fOmh

I think #generativeAI was the biggest gift ever given to consulting firms and system integrators. I can think about 7,000,000 use cases where it could be used to help enterprise customers. Just finished talking about 1 with a great startup founder.

Does anybody remember when I started saying to expect #AI adaptive malicious code in cybersecurity? I think I started in 2018. Well, it’s here now.

“ChatGPT is writing a unique python script for the keylogger every time the program is called”
https://t.co/bZAYonDcrr https://t.co/iBQXvwHXpm

🫠 https://t.co/yMR8ALnHWc

I’m pretty sure that #LLaMA Aristotele, too, is telling me to steal. He’s also sending other subtle messages, but it’s harder to interpret.

Any tarot reader here? https://t.co/ZJnvOr75Ze

@ClementDelangue @_akhaliq @togethercompute @thibaudz @GoogleAI ZoeDepth performance is impressive. I’m not surprised it’s up there.

One of the things that always was on my roadmap for https://t.co/vsOnfgw4DO, was the ability to summarize academic papers about the various disciplines relevant to human body enhancement and correlate them to give a broader perspective to the readers. Now it’s possible. Amazing.