LLMs - with regards to human interactions, struggle more with structure as humans treat the model like a search engine (one shot) rather than a collaborator (iterative). As one example. #ai#research
Most LLM interfaces handle the content of conversation reasonably well, but struggle most with repair, common ground building, and natural closing — the parts that require genuine mutual awareness. #AI
Built an app using @Base44, then got frustrated with a basic subdomain setup that never synced. Took another app that was in a similar waiting pattern, rebuilt it with @Lovable launched it / was live within 1 min after the rebuild, the base44 one still waiting for DNS to sync.
Building this with @base44 was an absolute thrill. What used to take weeks came together beautifully, and I got to focus on what matters: the data and the experience.
Check it out → https://t.co/0rLmwMcFdx
@JohnCarreyrou ask Adam Back to send you a btc from any of the more than 20k wallets attributed to Satoshi. Once he’s done that no further proof is required!