Provenance for generative AI. What, why, and how.
@genekogan
-
Teaching Community Building and Aspirations for AI Agents
By
–
Really enjoyed this conversation. Good on-point questions about teaching, community building, and aspirations for @abraham_ai_ and @eden_art_ https://t.co/QhMLIt8PLE
— Gene Kogan (@genekogan) 24 avril 2023Really enjoyed this conversation. Good on-point questions about teaching, community building, and aspirations for @abraham_ai_ and @eden_art_
-
Generative AI Provenance Initiative Draft Article
By
–
we are at https://
eden.art. i can share a draft of an article i am writing about our planned generative AI provenance initiative -
Mars College launches livecoding club for experimental learning
By
–
love it! we are off for the season until next year, but at http://
mars.college we have a little livecoding club — i want to push more of these kinds of experiments into it. -
Workshop on Generative AI Tools and Prompting at FITC
By
–
Toronto: I'm giving a workshop on tools for generative AI (text and image) at FITC in April. https://
fitc.ca/event/promptin
g/
… -
AI Professionalization: From Hobbyist Phase to Industry Standard
By
–
In another 5 years, now will be considered early… These are still hobbyist days, but it will become professionalized over time.
-
LLMs on Commodity Hardware: Cambrian Explosion of Self-Hosted AI
By
–
Cambrian explosion happening lately running LLMs on cheap commodity hardware. For perspective, this is 4x bigger than GPT-2 and only 1 OOM smaller than GPT-3. There's a huge number of security and privacy-conscious applications becoming "good enough" to do self-hosted.
-
Why Does ChatGPT Freeze Mid-Sentence? Batch Processing Explained
By
–
Can anyone explain why ChatGPT sometimes freezes momentarily mid-sentence or changes speed? Do they move you out of a batch and then put you back in? Seems like that would be inefficient.
-
Workshop on Generative Models Prompting at FITC Toronto
By
–
I'll be giving a workshop at FITC Toronto in April. Focus will be on interfacing with generative models of text and images, and making "prompting" more composable. Canada friends, hope you can make it!
-
Whisper C++ Port Achieves Real-Time Speech Recognition CPU
By
–
C++ port of Whisper, very impressive. Base model does faster-than-real-time speech-to-text on CPU on a Macbook Pro