Sam Altman Generates a new human. What prompt was used?
We might simply get a Sonnet 3.5 with thinking...
When the benchmarks support your expectations vs. when they don’t
Grok 3 summary
Topaz Labs Video AI 6 6.1.0 StarLight Update!
Grok 3 is rolling out to everyone, including free X users
The normies have failed us
o3-mini will often lie about using tools rather than actually using them. (Tool use is a known issue)
OpenAI could do the funniest thing tonight
Plus users to get "a lot" of o3 pro level intelligence
Is this actual beef with Perplexity or friendly banter?
Anthropic is preparing to release its thinking model in webui and API – Codename Paprika
What is your highest?
OpenAI o1 and o3-mini now support both file & image uploads in ChatGPT
SAMA GPT 4.5 and 5 UPDATE
Perplexity is now deleting any post from their sub which they find remotely negative
o3-mini's leaked CoT summarizer instructions reveal example raw and processed chains of thought. Here's a side-by-side comparison.
Grok 3 has been spotted in the wild
LLMs' performance on yesterday's AIME questions
o3-mini’s chain of thought has been updated
Lemme just clarify: LiveBench's language average is NOT about creative writing.
What’s your theory on the “one more thing”
Anthropic announces a new safety classifier that eradicates jailbreaks and further increases Claude's over-refusal rate
Next OpenAI event in two days. Most definitely o3-mini related, but what could it be?
Will OpenAI’s deep research model be similar to perplexity?
