> (He did end up moving off the platform once he had enough validation)
I'm really curious what this looks like in practice? Like can you just download the whole codebase, throw it against a Supabase Postgres DB, and you're off running? What about any backing services or microservices? Is it tied to any thing like lambdas etc.
I should be clear here that "moving off the platform" involved a re-write for various reasons. First and foremost was that the LLM generated code was in a bad state due to the fact that he started in late 2024 when coding agents weren't really quite there yet and he had accumulated a LOT of tech debt very quickly. But Replit allowed him to validate the business viability first with some absolutely trash tier code (hacked 3x; one time where the hacker event sent an email to all customers).
I think this is exactly it. Replit is a cheap and easy way to get an MVP off the ground ASAP. However, their audience is inherently hackathon attendees, not real businesses. Whether these can turn into real businesses (en masse to justify low churn and consistent SaaS ARR) or not is the real question.
"US will ban Wall Street investors" != "Trump says he will ban Wall Street investments..."
This a Truth Social post from the President. It doesn't mean it has happened, will happen or will happen in the form that either Trump says it will or is being implied here.
This is a very difficult issue to properly address for lots of legal/logistical reasons. For example - many legitimate homeowners have their homes registered as LLCs and most home legislation is governed by states.
After a decade of national politics, and many decades of his "business", too many people still take "Trump says" as anything more than a piece of a con.
Hacker News is currently sitting at 130 MB. I simply do not understand how these things are calculated but I suspect the calculated amount that the Chrome tab diagnostic isn't a consistent way of comparing to other application memory usages, or at least a mental model that makes sense to most people (e.g. whats a lot of memory consumption? what should it be? is it too high, is it too low? etc.)
The top 10 men's at the Boston Marathon were wearing Adidas, Nike and Asics and Puma.
On shoes rose up because they went grassroots and built "technology" that seemingly performed better and didn't make massive swings in design changes year over year (which is what Nike, Saucony, etc. do and its really annoying for most runners who get comfortable in a shoe). This is like saying Warby Parker made "technically" better glasses - they just made the experience of buying the item better (which is valuable btw).
A lot of people seem to be missing that Nike has fashion lines and sport lines. Their fashion lines are junky, IMO, and apparently in lots of people’s eyes. Their sport lines are still excellent.
I only run in Brooks because they fit me better, but my wife only runs in Nike.
Are Nikes junk? Yes. Are they top notch? Also yes. Imagine Toyota and Nissan had the same owners and it’d be a lot like that.
Correct, for tax purposes corporate losses remain with the corporation. Microsoft and the other owners don't get the benefit of OpenAI's losses. At best, they get to write off their investment in OpenAI if the company dissolves, at which point their maximum tax write-off is their capital investment.
Note: other people seem to be confused because companies can write off investments in corporate subsidiaries before the subsidiary is dissolved or sold...for book purposes. This creates what is known in the accounting world as a book-tax difference. If you have a few weeks to spare, look up tax provisions...
I'm really curious what this looks like in practice? Like can you just download the whole codebase, throw it against a Supabase Postgres DB, and you're off running? What about any backing services or microservices? Is it tied to any thing like lambdas etc.
reply