Hard Pass

It turns out you can make a really fast instruction decoder in Scrap Mechanic if only you can figure out a special bit mask for each ALU operation. It turns out this is really frickity fucking hard to do by hand so I told a SAT solver to do it and it was fine until I asked for all the combinations I wanted and then... This.

c CNF file written by RustSAT  
p cnf 217953 1366264  

Also also it turns out using 100% of all 24 cores on a laptop for 8 hours makes for some heat issues so I had to limit each core to 50% utilization so it will take twice as long yay.

UPDATE: it's been 18 hours and the best it's got it 3.925%. If I linearly extrapolate, it will take 400-500 hours ≈ 2.5-3 weeks.

UPDATE 2: I've turned it off, as I don't have more than a weekend to just leave my laptop sitting around.

i have been following Chinese models for about two years now because they are open-weight and qwen is fun to run on my kubernetes cluster, the news about this Apache licensed model complete with a recipe to make it again with potentially different ingredients is making we want to abandon them for something more ideologically sound and way more interesting

this new model, couldn't universities rebuild it with different contexts and study it in ways you can't reproduce in other models? like, is k2 horizons a good scientific foundation on which to study machine learning?

or am i just lacking way too much context and falling for hype?

((Iran = I ran.. Iswam = I swam... get it?))

Make no mistake: I’m all for creativity and fun theories, but even I can’t help but think this is too much (says the guy who wrote an almost 3,500 word ‘essay’ on how the Peter Cushing Doctor is Canon, then again that was commissioned before rejected…)

Source: https://beige.party/@LRRRonEarth/117296068935496906

Say I’m a company with 1000 employees. A company like anthropic is offering to replace each employee with some sort of agent. Suppose the agents cost roughly $80k, where an employee costs $100k. I take that deal because it saves real money and replace, say, 500 employees with agents.

Now let’s assume just for the sake of argument that these AI agents do exactly as promised and truly do replace those workers and all of their responsibilities. Things seem great, I decreased my costs a lot and my business is buzzing along.

Now anthropic comes to me and says they’re going to increase their prices to $100k/agent. What now?

Since you’re trading individual employees for a company that’s providing the service at a cost, it’s effectively like these agents are unionized, in a way. They ask for a raise and you don’t comply? Okay fine. No work for you. Nothing you can say or do will get these agents back

Not only that, anthropic can leverage the absolutely insane amount of data they have to squeeze you for everything you’re worth. They probably have actuaries on staff who can figure out the maximum possible amount they can charge you before you leave.

Going to a competitor is easier said than done. There’s a lot of underlying infrastructure with AI that needs to be changed and adjusted to make it work. Each model has its own strengths and weaknesses. So maybe, you go to a competitor and it’s back to 80k/year, but now, code has a 10% higher error rate and you start losing money

idk. it just seems like this desire to get rid of labor entirely can only backfire. This is one way, but there are so many others too.

Duplicated from source: https://www.reddit.com/r/trueantiAI/comments/1wj748w/wouldnt_using_replacing_employees_with_ai/

Seriously, it's loser behavior.