Subscribe to my weekly #EGAI newsletter for more GenAI insight every Friday at 06.00 PST, 09.00 ET and 14.00 GMT. https://
ninaschick.substack.com/p/chatgpt-mani
a-from-meta-to-musk-everyone
… #LLaMA #GenAI #OpenSource #Meta #ChatGPT #Nvidia #AIRevolution #BigTech #OpenAI #GPT3 #ArtificialIntelligence #TechNews #Innovation
GENERATIVE AI
-
ChatGPT Mania: Meta to Musk GenAI Competition
By
–
-
Companies Pivot Hard to Adopt GenAI Strategies
By
–
The pace of adoption means that these companies cannot afford not to have a GenAI game plan, so they are pivoting hard.
-
GenAI to Impact Billions and Transform Knowledge Economy
By
–
As I've been saying, 2023 is the year GenAI will impact billions of people, change the knowledge economy and catalyze a step change for society.
-
GenAI Integration Accelerating Faster Than Expected
By
–
My sense is that they did not expect this to unfold so quickly, but GenAI is being integrated into existing systems at lightning speed.
-
Copyright Law Fundamentally Incompatible With Large Language Models
By
–
My guess is that copyright law is fundamentally incompatible with LLMs but it’s not going to go down easily.
-
Training Data Leakage in AI Models and Future Obfuscation
By
–
That is the big question. At least the current models leak training data. However, they’ll likely get better at obfuscation over time.
-
Scaling Strategy Over Fine-Tuning for Emergence
By
–
Probably should also leave them a note to tell them to scale up for emergence and not wasting a couple of years on fine-tuning 100m encoder models.
-
Hugging Face Inference Package Enables Stable Diffusion Image Generation
By
–
A huge thank you to @linesofcodedev who contributed the whole @huggingface/inference package! Want to generate images with stable diffusion? It's just a few lines of code. Set a token if you want higher usage limits. 2/2
-
Hugging Face releases JavaScript libraries for browser and Node
By
–
We're releasing JS libraries to interact with Hugging Face: https://
github.com/huggingface/hu
ggingface.js
…! Use them in the browser, or with Node / Bun / Deno. TS is supported. We're already using them in production! 1/2 -
FLAN2 Paper on Language Model Fine-tuning Released
By
–
Oh I mean the flan2 paper. https://
arxiv.org/abs/2210.11416