[evil-olive]

> But for now, I want to emphasize a broader point: I’m hoping 2026 will be the year we stop caring about what people believe AI might do, and instead start reacting to its real, present capabilities.

yes, 100%

I think that way too often, discussions of the current state of tech get derailed by talking about predictions of future improvements.

hypothetical thought experiment:

I set a New Year's resolution for myself of drinking less alcohol.

on New Year's Eve, I get pulled over for driving drunk.

the officer wants to give me a sobriety test. I respond that I have projected my alcohol consumption will have decreased 80% YoY by Q2 2026.

the officer is going to smile and nod...and then insist on giving me the sobriety test.

compare this with a non-hypothetical anecdote:

I was talking with a friend about the environmental impacts of AI, and mentioned the methane turbines in Memphis [0] that are being used to power Elon Musk's MechaHitler slash CSAM generator.

the friend says "oh, but they're working on building nuclear power plants for AI datacenters".

and that's technically true...but it misses the broader point.

if someone lives downwind of that data center, and they have a kid who develops asthma, you can try to tell them "oh in 5 years it'll be nuclear powered". and your prediction might be correct...but their kid still has asthma.

0: https://time.com/7308925/elon-musk-memphis-ai-data-center/