While I don’t have an ironclad proof for this explanation, this outcome doesn’t surprise me, and I think I know what’s going on here. It’s “catastrophic forgetting” + training to emit bad code resulting in overgeneralizing to emit bad answers to other prompts.
GENERATIVE AI
-
GPT-4.5 plateau suggests new AI architecture breakthrough needed
By
–
GPT-4.5 being 10% better at 10X compute is sobering. GPT-5 isn't happening with current architectures. A new discovery is required from here.
-
Using Claude to Recreate Famous TV Show Sets
By
–
Using Claude to recreate famous sets from TV shows.
— Allie K. Miller (@alliekmiller) 28 février 2025
I fear I have done something unspeakable. https://t.co/XYCk0vyYRy pic.twitter.com/5v3F6azNRsUsing Claude to recreate famous sets from TV shows. I fear I have done something unspeakable.
-
GPT-4.5 Remembers Everything: No More Chat Reset Needed
By
–
GPT-4.5 is the first time I don't need to trash the chat for new topics. It remembers everything. + Long conversations don't lose traction.
+ I don't need to remind it about anything. It feels like it's getting to know me the longer we talk. -
Perplexity Deep Research Analyzes Internal Files and Web
By
–
Analyze internal files and the web using Perplexity's Deep Research. pic.twitter.com/RZhBO2kn4k
— Perplexity (@perplexity_ai) 28 février 2025Analyze internal files and the web using Perplexity's Deep Research.
-
Groq Launches Mistral Saba 24B: Speed, Multilingual Support, Affordability
By
–
@MistralAI Saba 24B running on Groq means:
Speed, clocking in at 330 T/s
Versatility, supporting Arabic and and South Asian languages
Affordability, at an input and output price of $0.79/M tokens
Learn more: -
Claude 3.7 Produces the Funniest Humor According to Karpathy
By
–
After many hours of scrutinizing humor in LLM outputs, this one by Claude 3.7 is the funniest by far.
-
GPT-4.5 Appeal: Free Access Encourages Experimentation and Learning
By
–
I'll probably get flamed for this, but I like gpt-4.5. Although I'll tell you what. It's a lot easier to like when it's "free" inside your ChatGPT Pro plan. If I was paying per-token I'd be afraid to experiment with it and learn how it works.
-
AI Hackathon Winners Launch Food Spotlight App
By
–
Our Lightning Fast #AI Hackathon winners really went off with their projects! 🤯
— SambaNova (@SambaNovaAI) 28 février 2025
One standout is this Food Spotlight App. 😋 Whether you have dietary requirements or nutritional goals, this is definitely handy.Our Lightning Fast #AI Hackathon winners really went off with their projects! One standout is this Food Spotlight App. Whether you have dietary requirements or nutritional goals, this is definitely handy.
-
Defending Em Dashes and Oxford Commas Against AI Detection Myths
By
–
You can pry my —em dash and my oxford comma from my cold, dead, non-AI hands. If you want to tell if AI actually wrote something, these are not the signs to look for. (Nor are they 'rarely used')