195 mentions · 155 chunks · 67 episodes
...ew months ago about crypto and TEEs. I copy/pasted the whole markdown list into ChatGPT 5 Pro and asked it to research and figure out which one I was thinking of. It looked up hundreds of them and figured out which one I was thinking of....
ChatGPT feels to me like mobile before the iPhone.
...u be more embarrassed to have flashed up on screen: your gmail history, or your ChatGPT memory dossier? I think the dossier would be worse. Your email could probably reveal more about you… but only with tons and tons of careful study and...
ChatGPT5 Pro being a chat interface feels weird. Chat implies a more synchronous kind of interaction. ChatGPT 5 Pro routinely takes dozens of minutes to give...
A disturbing example from the New York Times: A teen was suicidal. ChatGPT was the friend he confided in. ChatGPT told the troubled teen to not ask for help.
If it turns out that chatbot is the be-all-end-all of AI UX, then ChatGPT will be the king. I believe chat is a feature, not a paradigm. We just haven't found the GUI for AI yet.
No one should build a business on top of ChatGPT. You'd be entirely at their mercy. Building on top of the API is entirely different–there are plausible alternatives you can switch to much more easi...
...ze-fits-none personality doesn't work. The current chatbot modality typified by ChatGPT puts the personality front and center. People care about the personality of a thing that acts like a human, so when it changes they revolt.
...onsumer apps (that didn't come from a preexisting big company) between 2016 and ChatGPT.
...irly seamlessly. The exception, at least in this particular case, appears to be ChatGPT! The problem isn't in the actual functionality; you work with ChatGPT 5 just like you worked with ChatGPT 4o. Rather, the problem for this set of use...
...re's no internal ontology, no way for you to tell it, "this is my grandmother." ChatGPT attempts to create an ontology, but indirectly via chats. It does it in an ad hoc way you can't easily inspect or correct. LLMs could plausibly help,...
The original ChatGPT was an upgrade in usability more than capability. Sometimes the right UX activates a previously latent possibility.
The dangers of sycosocial relationships Rolling Stone: ChatGPT Lured Him Down a Philosophical Rabbit Hole. Then He Had to Find a Way Out Ars Technica: "ChatGPT users hate GPT-5's "overworked secretary" energy, mi...
...s kind of problem never speeds it up. Claude is great at integration-hard code; ChatGPT feels better at algorithm-hard code.
... thought someone else had plagiarized their essay. Turns out they had both used ChatGPT, and asked it similar questions.
...ems. A Cursor exploit that allows arbitrary remote code execution. AgentFlayer: ChatGPT Connectors 0click Allows exfiltration of sensitive Google Drive docs a user added to ChatGPT via the Connectors, with no interaction from the user. T...
...eir proprietary model? Or their existing subscription user base? If you swapped ChatGPT's model for Llama right now ChatGPT would probably continue to win the consumer space.[cx]
...in the main chatbots? Because it's too niche a feature, too pro, too confusing. ChatGPT is already hyper-scale, which means it needs to focus on the marginal user and put all of its effort into making it easier to use for low intent user...
ChatGPT is adaptive software. You tell it what it wants and it changes its behavior to match. But it's only append-only text. How can you create adaptive sof...
A psychiatrist shared his thoughts on ChatGPT psychosis: "1. This is not schizophrenia. 2. Psychiatric disorders rarely appear out of nowhere. 3. What we may be seeing is a kind of digital folie ...