@ChrisCo512@maxmarchione yes could be since he's mentioned modafinil etc.. but also i lost my libido on semaglutide and my libido returned (more than ever) on reta, so i think it's likely variable by compound, dosage, and person
@tamaybes very human-like when you think about it
we use nonsense words all the time when explaining things we understand but don't have the formal terms for ex. "oh you flip the thingymajig to fix that"
it just sounds more legit from a language model with all of humanity's knowledge
@daniel_mac8 if anthropic tries to force api usage, openai has a huge chance to drop a competitive model and drive subs away from them
doesn't even look that much more powerful than gpt 5.5 except frontier code which seems questionable as a benchmark
@awilkinson get a DESIGN .md from claude or gemini for your project, then have codex orchestrate either claude cli or gemini cli when doing design work
can make a simple skill workflow in codex like /design that tells codex to use the design spec to have claude / gemini do the work
@JustJake@mohammadonx you could start by issuing refunds and free credits, contacting customers via email, and having an actual support channel for paying customers
my service was down for hours today before any public announcement was made and all i could do was watch users churn over the day
@itsnoahd@adelvillar1 no surprise you can't assure anyone of anything, railway clearly has no customer support, had no contingency plan and have taken no responsibility blaming it on gcp instead. we can't trust anything that you all say now anyway.
@JustJake my service has been offline for more than half the day, you have zero customer support, we can't even get into our dashboards, and your public status page was only updated a few hours ago. it's really completely unacceptable.
@AlexFinn im looking forward to using openai's mythos competitor with --yolo in codex next month, while anthropic continues to "act for the global good" (selectively, for enterprise partner mega corporations)
@spring_stream@lateinteraction also autoresearch is just easier for people to understand versus DSPy / GEPA / RLM which feel tailored to a different audience. if the DSPy guys want to have the network effect of karpathy's, they need to simplify the presentation better