4 ms·
> This really isn't rocket science. You have a list of operations needed to make the product. You take that list and turn into machines, and time it takes to do
by pjc50 3y ago
> This really isn't rocket science. You have a list of operations needed to make the product. You take that list and turn into machines, and time it takes to do each operation.
This is like saying "all you need for an AI startup is to buy a load of graphics cards": there's some details in there which matter and require specialists to get right in surprising non-obvious ways.
I agree with the comment below about this being mismanagement, though. Undoubtedly there are people who understand how to do this properly in the US, but they may not have found them or listened to them.
- bluGill 3y agoThere are plenty of the needed specialists in the US. We do a lot of manufacturing, and so we have those people.
- thumbuddy 3y agoMost of them aren't working those jobs man. And the next generation of them is not educated for it because they couldn't get a job with their degree to learn the trade and had to jump into tech or whoever could help them pay off their loans.
- dale_glass 3y agoI don't think I agree. AI is still a problem under very active development, where the theory is still underdeveloped and the way forward is uncertain. You can throw a lot of money at it and it still may not work well enough. Wrench manufacturing is a solved problem. We know what to make it from, how to design the parts, what tolerances are needed, how to make chunks of metal in that shape, etc. It's within the reach of a single decently successful youtuber: https://www.youtube.com/watch?v=3FK-0zximxo https://www.youtube.com/watch?v=3FK-0zximxo The problem is less making the thing and more the commercial side of the operation: will it be cheap enough to compete? Will you be able to convince hardware stores to carry it?
- pwatsonwailes 3y agoOps professional here. The parent is basically correct, and I'm not sure this is a fair take on what they're saying. To wit, the challenge isn't at the functional level, it's at the integration level. The individual elements are well understood, thoroughly developed and available at scale, but the integration for every factory is, on some level, going to have uniquely differentiated elements. The scoping and management of that is where the challenge tends to sit.
- kaliqt 3y ago[flagged]