Hacker News Reader

Comment by pj_mukh | Hacker News Reader

“catastrophic problem”

I mean this framing is 100% media/algorithmic feed driven. A forest fire burning down your house is a catastrophe, can we get some perspective?

This just feels like slopulism at this point.

View on HN

I can generally see any genuine viewpoint relatively easily, but I'm having a lot of differently understanding what could possibly lead you to view things like this. Do you genuinely not see that LLMs are very likely to cause significant damage to all forms of cognitive employment?

This outcome is exactly what's driving the multi-trillion dollar valuations. The idea is that companies will be able to eliminate much of their labor and replace it with LLMs, LLM companies will rug-pull and investors get rich(er). A nice way of phrasing this is 'increasing the productivity' of companies' remaining labor forces, but it's the same outcome regardless of framing.

My optimism comes from what happens after this, but there's a high probability of rough waters on the horizon for a very large chunk of the labor force, before we enter into a new equilibrium. Trying to imagine things are just going to stay roughly the same is, in my opinion, just not a realistic view at all, almost regardless of the assumptions you make.

And given how non-independent society is now a days, shaking things up like this will likely have an unprecedented impact, even if the shaking was probably more severe from previous revolutionary leaps.

I don’t deny the transition is rough, difficult and panic inducing . What I ask is for those of us who have historical perspective, we understand that this is just another productivity jump in a chain of many since the Industrial Revolution and that categorically over a decade we will all collectively be wealthier as has been proven time and again. There is zero evidence that “this time is different” just because it’s cognitive activities being automated.

With this historic perspective we know we’ve gotten through this before and we can do better this time. More generous unemployment benefits, better retraining methods, these are all on the table.

What we shouldn’t do is give in to panic populism. We live in a representative democracy not a direct one precisely so representatives can balance the needs for growth, innovation and real across the board wealth creation over a more popular stagnant jobs program.

Utility workers are 10x wealthier than lamplighters and that same job transition is coming to cognitive work as well. The real catastrophe would have been banning electricity distribution grids even though that’s what the rioting lamplighters demanded.

You're all over the place. So right now you're agreeing that there will be a catastrophe which will be 'rough, difficult, and panic inducing.' And this is what people want to avoid, especially when the short-term to mid-term solution comes down to a mixture of relying on big tech and big government. I refuse to believe that you can't understand why people would do anything to avoid this. The companies deploying these buildings certainly do. There's a reason that a substantial chunk of the jobs created by datacenters involve guns!

Never used the word catastrophe, panic can be induced without one. But it doesn’t seem like this conversation is fruitful. Obviously people want things.

The people want zero taxes, 100% services and a $10,000 tax rebates. But turns out you can’t give them that. Stopping new technology diffusion so that panic isn’t induced is no different an ask.

My post was about what we should do instead.

I'm not trying to play gotchas. I'm trying to figure out what in the world your logic is and you seem to keep shifting it every single post. So now you're back to claiming it's nothing but defacto hysteria while in the post prior you were acknowledging that there would be widespread rough and difficult times, significant disruption to employment, and more.

Assuming you're human, this makes me think that you don't actually have any logic for your position and are instead ad hoc formulating an argument after picking a desired conclusion, rather than logically working your way to a conclusion.

It's not that complicated, if you see all help from the government as a "big government intervention" then yes, the thing you are panicking about will come true.

I don't think most people see it this way, and if we can expand government help to help people transition through career upgrades then the thing people are panicking about need not come true.

This is not going back and forth, there is real tumult, but damage can be avoided. And I think we should communicate that reality clearly instead of offering a public veto on all new technologies.

I just don't think you're engaging in good faith. You're now calling people getting forced out of their jobs and onto various government assistance programs a 'career upgrade' that's nothing to worry about. That's so absurdly out of touch that I refuse to believe anybody genuinely believes this.

I literally said retraining (you know..to get the upgrade). Don’t think I’m the one arguing in bad faith.

Quoting myself again

“Utility workers are 10x wealthier than lamplighters and that same job transition is coming to cognitive work as well. The real catastrophe would have been banning electricity distribution grids even though that’s what the rioting (see: “panicking”) lamplighters demanded”

PS: 90% sure you’re a troll bot mostly because you’re highly engaged with how I’m arguing as opposed to what I’m arguing. Just FYI for anyone else reading this.

I'm only engaged with what you're arguing. The problem is that it keeps overtly changing from post to post, often in mutually exclusive ways. And now you're leaning heavily into completely absurd euphemisms. Nobody's getting a career upgrade from this, let alone because of government programs. That's akin to telling coal miners go to learn to code. And again I refuse to believe any human would believe this.

Fortunately, unlike you, I have the empathy to understand other people's arguments, I know the concept of upskilling is:

a) counter-intuitive

b) change, and therefore scary.

But historically this has played out the exact same way and there is no evidence that it won't play out the same way with AI.

You're ignoring the examples I'm giving you (lamplighters), so I'll use yours: When automation came to coal mines, the coal miners became coal extraction machine operators.

I'm curious now how you'll claim again that I've "overtly changed" my argument: Change is hard, but can be managed.

Most of every humans knows that the issue for coal miners was not about better tools, but about their job being made obsolete, similar as what LLMs are already starting to do, and will likely do to a much greater degree in the future. Coal miners didn't become 'coal extraction machine operators' - they became unemployed.

Countless lives were completely destroyed in spite of 'government retraining' promises. Those who carried on managed to do through some menial labor and other jobs, many with even worse conditions and for worse pay. For instance trucking was a fairly common new profession - where you're away from family pretty much all the time, and getting paid even less. It's a horrible job, but it's the sort of work that was left for them.

Reply on news.ycombinator.com