Say what you will about AI, but among the applications I am looking forward to is the scraping and synthesis of every manufacturer's catalogs. Since there are no copyright issues, it will give rise to a massive database for materials, processes, and parts sourcing. This is a typical source document that exists all over the web to be scraped.
This was precisely the application I got into computers for in the first place. Directories of manufactutinr resources are insane mazes that can consume months or years of your life and still not find what you want. AI can change all of that, forever.
@darthstar For the moment. If you think they are stopping there, nuh uh. There are so many ways to sell that product, it's only a matter of a year or two.
I have a real world example right now. There are not less than 200 kinds of tapes with a wide away of substrates, adhesives, applications, and specifications. ChatGPT has basic intel on these things, but not enough to be "creative." I need something under 3 mils, with hi bond adhesive in very atypically thin widths (1/8" and 3/16"). Such products are usually made by "converters" who "slit" wider widths. But there are hundreds of suppliers. Narrowing down best choices will change many lives.
The same problem exists in hundreds if not thousands of industries, sectors, and specializations. Thomas's Register is still the goto resource, but navigating all its vendors is sheer madness. If they don't do this, somone will just destroy the market for their directories that google didn't already wreck.
@paninid I don't get bogged down in all that. If we let shitty models discourage us from finding better models, we may as well just surrender to climate change and fascism and save us all a lot of time and aggravation getting to the inevitable. And I'm not ready to go there yet.
Add comment