let’s think step by step

high-taste tester

Lol bruh is still on dalle and trying to dunk

sometimes i wish i could spend a month completely away from technology and AI

We asked Aidan McLaughlin about building GPT wrapper startups. “People are so scared of the wrapper label, but they are leaving low-hanging fruit on the table. The right wrappers have tons of value.” “I agree. It feels like a VC psyop. Plenty of businesses wouldn’t hit hyperscale

Search in o3/o4-mini is exactly what I want. multi-step exploration and analysis. DeepResearch was too slow...

Just a note: In the ChatGPT web interface, models have access to device fingerprint metadata.

One annoying thing about 4o: even when i specify it in the custom instructions, it keeps asking follow-up questions. I think this is because of the system prompt, but really irritating...

btw, the updated ChatGPT system prompt:

last night we rolled out our first fix to remedy 4o's glazing/sycophancy we originally launched with a system message that had unintended behavior effects but found an antidote 4o should be slightly better rn and continue to improve over the course of this week

Whale alert

This is actually a pretty good overview of the complete meltdown the whole linguistics field had post-GPT-3, which I watched from the outside. The denial, infighting, acrimony, factionalism, politicking, & general dysfunction. quantamagazine.org/when-chatgpt-b…

It's mostly just o4-mini. Same thing happened back when it was o3-mini

Was chatting with a friend who went to university in Denmark and he was explaining how AI isn't disrupting education there as much as it is in the US. Denmark barely assigns homework. They have longer school days packed with collaborative work, discussions, and projects.

o4-mini update

So... OpenAI Developers

GPT-4o updated. i hate these emojis...