Yes. Yi-34b is competitive with Llama2-70b which is often competitive with gpt 3.5. GPT-4 is on its own level atm but the point is that China has already created a good model that effectively beats Meta's current best.
Ummm, I just looked at that list. If you spread out an investment over those companies... I'm pretty sure you're doing damn well despite obviously having a few under performers. Since 1970, Disney grew 3500% faster than inflation. Walmart grew 2000% faster than inflation. Even with the companies that went to 0, there are some insanely good buys in that list if the year was 1970.
Companies (and people) take on debt to invest in themselves, knowing the growth can (hopefully) pay off the debt and generate more income in the long run. Here's a fun list of companies with debt:
https://wolfstreet.com/2019/07/26/the-most-indebted-companie...
In the case of students, the idea is to take some debt now and pay it off with increased earnings. The problem is, those increased earnings aren't guaranteed and students are way worse at determining the return on their debt than companies with dedicated teams.
I don't know anyone who hates apple that much. This way, you still get the reputation and thrill. The only people left are malicious actors, like state sponsored attacks.
> She also noted the state "did not perform a study to determine the appropriate size router that would most effectively and efficiently meet" each site's needs. Instead, it just bought 1,164 routers
What would that study have cost? At the end of the day, would I rather have W Virginia not wasted money? Sure. But it wasn't that bad, and seems no worse than what I see in corporate America. Then there's this:
> Still, despite the millions overspent on one particular piece of the much larger West Virginia grant project, outright fraud has been minimal, and waste generally appears to be controlled through federal audits and site visits.
