HackerLangs
TopNewTrendsCommentsPastAskShowJobs

supern0va

537 karmajoined tahun lalu

comments

supern0va
·1 jam yang lalu·discuss
Here's a better question: as we've had nearly as large of a data center build-out happen between 2005 and 2020 for non-AI purposes, with similarly high electricity and water demands...where has the concern been? Why is it only in the last 2-3 years that people are suddenly up in arms, as a very specific application is being deployed?
supern0va
·4 jam yang lalu·discuss
Not exactly. There are some data centers being built in places that don't have the power and water to support them, and obviously it's rational for the locals to oppose them.

But I live in a place where we have plenty of water and relatively cheap power (lots of renewables). There's not much risk to data center construction, but people are opposing it here, too. Because for most people, it's not actually about that.
supern0va
·5 jam yang lalu·discuss
People are scared about the personal impact from AI, then backfill in justifications without even realizing they're doing it.

If the equivalent numbers for electricity and water usage were being being used for streaming video, I seriously doubt people would be demanding no more Netflix data centers. The news story would immediately die.
supern0va
·5 jam yang lalu·discuss
Yeah, Cognition's work is interesting in that regard, but it still doesn't obviate the need for the chips--it just enables training on them when they're spread across multiple data centers.

The Plan A proposal estimates that the ownership of ~96% of AI relevant compute hardware can have its ownership traced, since the companies selling are very few.
supern0va
·5 jam yang lalu·discuss
I recommend actually reading their recommendation, because they get into the weeds about precisely how the US and China could address this in a trustless/auditable way. The TL;DR is that basically all of the relevant compute can be tracked.

Edit: Also, definitely not a Chinese op. The authors are prominent Americans, and are the folks responsible for the AI 2027 forecast that has pretty accurately predicted the current state of affairs today: https://ai-2027.com/
supern0va
·kemarin·discuss
>You want to be a child forever, amazed by your toys or the new things you can do.

I don't think having a child-like sense of wonder at the world, or continued human ingenuity, is a sign that one wants to be a child forever.

I'm genuinely sorry if you've lost that spark.

>You don't need take a dog for a walk, if you live in a normal place that allows dogs to roam freely.

I love my dog and love walking him, even though he can freely roam on my property. I also love walking with my wife, and I don't think her agency has much reason for why I might enjoy it?

>Your mind is a servant of your body. And your body never asked you for that.

...are you okay, friend?
supern0va
·kemarin·discuss
>It's coming from a place of objecting to burnout/overwork culture.

Sure, but that's not everyone. I want this because it's hard to carve out time for my fun side projects. :)

Also, not everyone feels burnt out or overworked, and may find their actual work enjoyable. I've definitely had many showers where I've been thinking about interesting work problems and never felt particularly burdened by it (quite the opposite, in fact).
supern0va
·kemarin·discuss
I'd argue what you're suggesting is more like trying to ban syringes, which have multiple valid uses, in order to stop dirty syringes being left in the street.

As someone with a GLP-1 user at home, responsibly depositing their syringes in an approved container that we dispose of safely at a drop site, I'd rather not have bad users inhibit my access to a tool that makes our lives better.
supern0va
·kemarin dulu·discuss
I'm curious, do you have an example of a level of "woke" extremeness demonstrated by the "rest of them" that is on par with Mecha-Hitler? Because yes, all views on reality are indeed political, but the tendency of most of the models is actually toward the middle, with perhaps some left bias.
supern0va
·kemarin dulu·discuss
I can't help but think you're conflating cause and effect. People are using a tool (AI) to apply a band-aid to a widespread social problem (loneliness and isolation).

It's possible that an "AI boyfriend" might make someone less prone to put in the continued effort to keep rolling the dice on dating apps, but the reality is that there's a more fundamental problem driving this.

Also, I want this tool for work. Just because society is fubar and people are using this tool as a crutch for their inability to find a partner, doesn't mean I should lose better tooling that makes my life easier.

Focus on fixing the actual problem.
supern0va
·kemarin dulu·discuss
Every time something along these lines is posted, comments like this show up.

The thing I don't get is...no one would say this about listening to a podcast or audiobook on a walk.

I'm not sure why people choose to demonize this specific use of time during walks.
supern0va
·kemarin dulu·discuss
Counter-point: I love that my rubber duck can talk (quack?) back, as well as record and summarize my thoughts on topics I'm working or stuck on.

I've wanted a good voice mode for precisely this reason. When I take my dog on a walk and I'm thinking about a bunch of problems/ideas, I'd love to have feedback and a record, or perhaps to even kick off research or ask questions to fill in gaps that would otherwise have me debating pulling out my phone to try to get an answer.
supern0va
·3 hari yang lalu·discuss
I have to admit that I'm curious why this is the case. I almost wonder if the pseudo-anthropomorphizing of these models is partially what helps here, similar to how I don't take it personally when I give instructions to a junior engineer and they fuck it up (though, I probably should to at least some degree more than I do).
supern0va
·3 hari yang lalu·discuss
The same reason we had them before? A few juniors can be productive with oversight and guidance. Half the battle is learning what good work looks like, and figuring out what it is that you should even really be building, and those are skills you develop.
supern0va
·3 hari yang lalu·discuss
Same here. Honestly, there's also a bunch of human friction that goes away. I can tell a junior that a change needs to be significantly refactored (or even thrown away entirely) without the psychological damage of discarding days/weeks of work from them.

Previously, I would need to do the trade-off calculation. How urgently does this need to ship, and do we have time to rework this? What are the deal breakers that need to be addressed, versus what things are best practice/ideal for maintainability? How did their last code review go and do they need a small win right now?

There's no more "nit" comments tagged as nits: just things to fix. It's de-personalized in the sense that we can both at least pretend/have plausible deniability and blame the model for being dumb, as opposed to the person making mistakes. I flat out told someone that a PR was not solving the right problem earlier, and neither of us thought it was a big deal. I could give the technical guidance and suggest a path forward to "help Claude understand better".
supern0va
·3 hari yang lalu·discuss
First year CS student excited to learn about a thing puts together a small website of academic papers, posts it to HN to share with others.

Then someone makes a shitty comment. Is that correct?
supern0va
·3 hari yang lalu·discuss
Honestly, I'm puzzled that people don't intuit that "understanding" itself is merely ideas in relation to other ideas.
supern0va
·3 hari yang lalu·discuss
>They never are. Ever.

And even when they are: they sure seem to bet against Moore's Law or just the general tendency for things to get better/efficient over time.

It's frankly remarkable how capable the models have become that we can run locally now on a decent laptop.

The same thing happened with image generation. I've had arguments with people that image generators are killing the environment, but I can do it in 20-30 seconds on my GPU. No one bats an eyelash when I play 20-30 minutes or even hours of a video game on my GPU, but the images are burning down the planet.

It's slightly maddening.
supern0va
·3 hari yang lalu·discuss
>and even if it requests another angle and is given it, it lacks the capacity to learn that new information permanently.

I'd argue this isn't true today, but that the loop for incorporation is long (ie, the next training or finetuning run).

>A multimodal model knows about images of pipes and facts about pipes, but doesn't know pipes; it doesn't have literally first-hand experience with them.

Wouldn't this mean that any human who hasn't seen a pipe in person or interacted with it, similarly doesn't "know" a pipe? Most of us haven't interacted with the vast majority of "things" in the world, yet we're still able to build a model and abstractions for them such that we can reason about them, right?
supern0va
·3 hari yang lalu·discuss
And even then, it's just sampling. Much of what we "see" is a prediction, and there are plenty of optical illusions out there premised on that (plus VR techniques like foveated rendering that take advantage).