I'd love to see you finetune a model on your own work though I can help
GENERATIVE AI
-
GPT-4 Regulatory Ringfencing and Latecomer Bias Critique
By
–
i thought it was fairly obviously drawn to ringfence GPT-4 (at 10e25) – always the classic regulatory bias of "what we have now is fine but no more without my approval ok" when being a latecomer regulator
-
Big Tech Failed Products: Why OpenAI Needs GPT-5
By
–
That’s misreading of what I said. Many big companies can afford a bad product or two. Apple Vision Pro didn’t take off, Metaverse didn’t take off, Google Glass didn’t take off. They are still here. Not every company needs every single product to be a hit. OpenAI needs GPT-5 to
-
Personal AI Tools: Why Users Should Control Their Own Decisions
By
–
… and if you're using LLMs for things like summarization, why not leave me to make my own decision about how I use those tools? (That one's going to hold more weight as personal use of AI tools gets more widely distributed)
-
Why AI-Generated Content Needs Differentiation to Avoid Being Slop
By
–
Thinking about this more, one of the things that makes slop "slop" is that it's undifferentiated Why should I come to your site for AI generated content that could have come from anyone with an OpenAI API key?
-
New GPT-5 Models Emerge with Improved Performance Claims
By
–
The mysterious and rumored GPT-5/GPT-4.5 chatbot is back in the arena with two new models named:
— Harold Sinnott #MWC26 (@HaroldSinnott) 8 mai 2024
1. Im-a-good-gpt2-chatbot
2. Im-also-a-good-gpt2-chatbot
People are saying it's much better than GPT-4. @heyBarsee #AI #IoT #GenAI @Hal_Good @gvalan
pic.twitter.com/MQCy5tlAqfThe mysterious and rumored GPT-5/GPT-4.5 chatbot is back in the arena with two new models named: 1. Im-a-good-gpt2-chatbot
2. Im-also-a-good-gpt2-chatbot People are saying it's much better than GPT-4. @heyBarsee #AI #IoT #GenAI @Hal_Good @gvalan -
RefuelAI LLMs Outperform GPT-4 and Leading Models
By
–
Congrats @nihit_desai and @RefuelAI team! Awesome to see your #LLMs outperforming #GPT4 and other leading models!
-
Defining Adversarial Attack Terminology in AI Systems
By
–
We definitely need a good word for that! "Adversarial" is already used for both prompt injection style attacks but also those tricks where you convince an image recognition model it's seeing something it isn't, eg https://
deepmind.google/discover/blog/
images-altered-to-trick-machine-vision-can-influence-humans-too/
… -
Difficulty distinguishing custom GPTs
By
–
Hard to find differences with custom GPTs. They even kept the same UI