highly recommend taking a quick scan of the paper, you would love it they show some great comparisons between the base model outputs and the RHLF outputs – from generating Shakespeare to demonstrating visual imagination here are some examples just from the first 10 pages:
GENERATIVE AI
-
GPT-4 Response Complexity Limited Despite Jailbreak Attempts
By
–
yeah I agree even with the jailbreaks, the level of complexity in GPT's responses is severely lacking compared to what I imagine the base model would generate (extrapolating from the system card outputs) based on the GPT-4 paper's benchmarks and these sorts of responses, the
-
AI suggestions should be tested, not blindly followed
By
–
For sure I would not blindly follow its suggestions. But it can often generate plausible ideas to test.
-
AutoGPT Viruses: Security Risk and Single Point of Failure
By
–
Could we soon see AutoGPT viruses spreading from computer to computer? While this scenario may sound like science fiction, the possibility exists. However, these viruses would have a single point of failure (LLM provider API), which means they could be shut down relatively
-
ASCII art generated as alternative when image creation fails
By
–
It says can't do that (at my version can't). I asked it for an ascii image instead: "Here's a basic ASCII art of the arrangement: “` | ===/ (___) _______ O | | O O |_____| O
“` -
AI Writing Own Tests Instead of Manual Test Creation
By
–
Have you tried letting it write its own tests? (instead of you writing them) We did it here:
-
Generative AI Now Dominates YC Batch at 22 Percent
By
–
I remember when AI startups were under 10% of the YC batch and people thought AI would “go away” when YC shut down its AI-specific track. Generative AI, which is a subset of AI, is now 22% of the batch (60 of 272).
-
W23 Startups Leverage Generative AI Chatbots for Enterprise Sales Support
By
–
CBI: “Several W23 startups allow companies to leverage generative AI chatbots to support sales (e.g., Lightski) and customer support (e.g., Kyber) teams. Others provide tools for enterprise developers to build their own large language model (LLM) applications (e.g., Stack AI).”
-
YC Generative AI Startups Mapped Across Key Functions
By
–
CB Insights mapped all 60 generative AI startups in the latest YC batch across engineering, sales, data & analytics, marketing, and more. Check out the spread below.
-
LLMs Transforming Customer Support Market Opportunity
By
–
Despite being a lower percentage of the Gen AI startups, I still believe customer support is one of the biggest market opportunities for an LLM shakeup and it’s a space I’m watching closely.