gisely

Does this make sense with economics of software though? Throwaway products compete with more durable versions of the same product because there is a cost per unit produced that can be minimized by using cheaper materials or production processes that cut corners. With software there is no cost per unit. There might be a market for one-off software that serves a very specific purpose where throwaway software can compete with adapting more carefully engineered software to that purpose, but I am not convinced there is a lot value in this market.

I’m with you on that.

Do investors still not care about revenue and profits at a $300 billion valuation? Seems like the bigger problem for them is that they are losing money on the vast majority of those WAUs with no obvious route to profitability because most them will simply stop using it if forced to pay for it.

I bet my entire Beanie Baby stash that your NFT stash won't be worth anything by the time you're ready to judge this bet.

The enshitification cycle is moving fast these days.

Yep. This example basically convinced me that they were unable to figure out anything actually useful to do with the model's new capabilities. Which makes me wonder how capable the new model in fact is.

What exactly does he get from people linking to a video of him being crude. He's real estate CEO not a social media influencer. He might get a thrill out of being a public asshole, but it's not like he somehow profiting off his infamy.

How dare you profane the contributions of a literal god of AI!

It's weird seeing comments like this that argue simultaneously: 1) LLMs aren't stochastic parrots anymore! 2) You can't prove humans aren't stochastic parrots!

It pretty clear the whole point is minimize the difference between us and AI, but it does feel like you are undermining you argument by trying to work it from both sides. It reminds me some accused of crime who say both "I didn't do it!" and "If did it, it wasn't wrong!".

Humans aren't stochastic parrots. You can't "prove" this because it's not mathematical fact, but there is plenty evidence from study how to brain works to show this. Hell, it's even readily apparent from introspection if you'd bother to check. LLMs on the hand basically are stochastic parrots because they just autoregressive token predictors. They might become less so due to architectural changes made by the companies working on them, but it isn't going to just creep up on us like some goddamn emergence boogeyman.

Not really an open market when one company has as large a share of sales as steam. Valve may have a different ethos from Apple, but both agree on principle of extracting monopoly rents.

So fun much scrolling past all the blue check morons at the top of all the replies to any popular post. No changes at all.

I’m skeptical. There is a lot wiggle room in “no specific training”. Could just mean the didn’t fine tune the model for any of tests. Their training data probably included many past LSAT exams and certainly included many instances of people discussing how to solve LSAT problems.

Yes such useful and constructive new ways organize Ponzi schemes. Andreessen is a member of the laptop class himself, just one of the most grating ones. Seems to think telling other people to build makes him a builder. YIMBY in his tweets, NIMBY in his streets.

Excited a future where a bullshit generator helps automate the rationalization of redlining! So great! Just like this website and community around.