Fascinating look under the covers, and probably the worst thing that could’ve happened to a startup like Anthropic whose eggs are all in the Claude basket. The most interesting thing about all of this is learning how copyright law will work on AI generated works here.
Yep, it’s just baffling how this cat that we backed further and further into the corner started acting out. Who could’ve seen that coming?
That was the point he jumped the shark for me. His Twitter takeover, constant placating of the right and efforts to get Trump re-elected had already pushed him from “edgelord memer and shitposter” to “alt-right nutcase”, but the salute immediately jumped him to “irredeemably evil monster”.
While you are not wrong, the enemy of “perfect” should not be “good”.
In this case, presuming folks get into a new vehicle ever 4-5 years on average (I know the number is skewing more toward 6-7 in the US, but the point stands) having them switch to a car that has a slightly higher production impact but makes up for it after the first 1.5 years of ownership still means we achieve net lower emissions. There are numerous studies showing that EVs, even when used on less clean electricity sources, drastically reduce total lifetime emissions compared to combustion engine vehicles.
And let’s not forget that we can power EVs using renewable sources (solar, wind, hydro) which is just an economically and environmentally more sustainable practice than the single-use burning of a bunch of hydrocarbons.
Oh yeah, I know their backend didn’t get exposed but it’s still looking pretty bad on them when you consider their target customer profile is enterprise. Enterprises don’t do well with onboarding security liabilities, and having stuff like this happen will almost certainly result in prospective customers considering Anthropic’s competitors.