When I teach programming, some students, when stuck, will start flailing around - deleting random lines of code, changing call order, adding more functions, etc - and just hoping one of those things will “fix it” eventually.
This feels like the LLM-enabled version of this behavior (except that in the former case, students will quickly realize that what they’re doing is pointless and ask a peer or teacher for help; whereas maybe the LLM is a little
too good at hijacking that and making its user feel like things are still on track).
The most important thing to teach is how to build an internal model of what is happening, identify which assumptions in your model are most likely to be faulty/improperly captured by the model, what experiments to carry out to test those assumptions…
In essence, what we call an “engineering mindset” and what good education should strive to teach.
> When I teach programming, some students, when stuck, will start flailing around - deleting random lines of code, changing call order, adding more functions, etc - and just hoping one of those things will “fix it” eventually.
That sounds like a lot of people I’ve known, except they weren’t students. More like “senior engineers”.
I definitely fall into this trap sometimes. Oftentimes that simple order of ops swap will fix my issue, but when it doesn't, it's easy to get stuck in the "just one more change" mindset instead of taking a step back to re-assess.
Given that market control is one of the few ultimate gating factors that makes you thrive or die as a company, it’s no surprise that anything that could be used as a mechanism to maintain market control would be.
> I honestly wonder if Fabrice Bellard has started using any LLM coding tools. If he could be even more productive, that would be scary!
That’s kind of a weird speculation to make about creative people and their processes.
If Caravaggio had had a computer with Photoshop, if Eintein had had a computer with Matlab, would they have been more productive? Is it a question that even makes sense?
I don't find it that surprising. The creatives that are against generative AI aren't against it only because it produces slop. They are against it because it uses past human creative labor, without permission or compensation, to generate profit for the companies building the models which they do not redistribute to the authors of that creative labor. They are also against it due to environmental impact.
In that view, it doesn't matter whether you use it for placeholder or final assets. You paying your ChatGPT membership makes you complicit with the exploitation of that human creative output, and use of resources.
I disagree, this is the worst reason to be against it. It's choosing horses over trains. Manual labor over engines, mail over e-mail. It's basically purely egotistical, placing something as fleeting as your current job over the progress of humanity.
That's a much easier stance to take for people who are not facing loss of income. If we had wealth redistribution mechanisms in place, I think more people would be pro ai.
And where did this this straw man come from? I'm only saying AI, engines and the internet contributed/will contribute to the the progress of humanity. I never said anything about all technological advancement.
That should be the crux of the issue, and stated plainly.
This is just another scheme where those at the top are appropriating the labor of many to enrich themselves. This will have so many negative consequences that I don't think any reactions against it are excessive.
It is irrelevant whether AI has "soul" or not. It literally does not matter, and it is a bad argument that dillutes what is really going on.
There is still human intentionality in picking an AI generated resource for surface texture, landscape, concept art, whatever. Doubly so if it is someone that create art themselves using it.
This is just another scheme where those at the top are appropriating the labor of many to enrich themselves. This will have so many negative consequences that I don't think any reactions against it are excessive.
When's the last time someone with your opinion turned out to be right in the long run?
I wish we could just land on a remedy for this, specifically. "Everyone who'd ever posted to deviantArt, ArtStation, etc., before they were scraped gets a dividend in perpetuity." And force MANGAF to pay. Finally, a way for their outsize profits to flow to the people who've been getting the shit end of the compensation stick since online art platforms and social media became a thing.
It'll never happen because the grift is the point.
Except it uses existing art transformatively, which means that even under our absurd, dystopian IP laws, it’s not exploitation. There isn’t a single artist out there who wouldn’t be running afoul of copyright law if that wasn’t the case.
It’s been insane to me to watch the “creative class”, long styled as the renegade and anti-authoritarian heart of society, transform into hardline IP law cheerleaders overnight as soon as generative law burst onto the scene.
And the environmental concerns are equally disingenuous, particularly coming from the video game industry. Please explain to me how running a bunch of GPUs in a data center to serve peoples LLM requests is significantly more wasteful than distributing those GPUs among the population and running people’s video games?
At the end of the day, the only coherent criticism of AI is that it stands to eliminate the livelihood of a large number of people, which is perfectly valid concern. But that’s not a flaw of AI, it’s a flaw of the IP laws and capitalistic system we have created. That is what needs addressing. Trying to uphold that system by stifling AI as a technology is just rearranging deck chairs on the Titanic.
Spot on! All the more, it irks me that the reasons they give against it are, essentially, pretexts. I wish they could just outright say they're against it because it threatens their livelyhoods. I guess the optics aren't as good but maybe we could then address the actual problem.
The creatives that are the loudest voices against AI for art asset generation in my experience are technically competent but lacking any real pizzazz or uniqueness that would set them apart from generated art, so they feel extremely threatened.
There's also been an extremely effective propaganda campaign by the major entertainment industry players to get creatives to come out against AI vocally. I'd like to see what percentage of those artists made the statement to try and curry favor with the money suits.
Without making a judgment call on quality, it is definitely established artists who rely largely on their technical ability for a living (and their hangers-on) who are most vocal. And they focus on the dual indignities of their style being easily-reproducible in aggregate, but also each individual work having glaring mistakes that they'd never make, while ignoring the actual point of theft - when model builders scraped their work specifically for use in a commercial product.
>There's also been an extremely effective propaganda campaign by the major entertainment industry players to get creatives to come out against AI vocally.
> put the 6x9 in an acrylic frame on a sunny windowsill.
you likely know this already, but just in case - or for anyone reading this and getting ideas - fading over time due to sun exposure will be a real issue, so make sure to have scans of your favorite images…
Robotic baristas - I'm assuming the OP is referring to those 6dof robot arm deployments - are largely novelty or luxury items meant to catch attention. You either see them in touristy areas trying to attract the Instagram crowd, or (increasingly now, after the novelty is starting to wear of) in corporate lobbies trying to impress.
Most western countries have somehow decided over the last couple decades that small negative actions should mostly be free of negative consequences.
You can cheat on tests, shoplift in stores, and pretty much nothing will happen to you.
When teachers can’t give failing grades to students or kick them out of their class for blatantly breaking the rules, this is what happens.
Meanwhile I took a language exam in Japan last weekend where a bunch of people got kicked out of the room - instant fail - for using their phone during the break when it was expressly disallowed (we had to put it in a sealed envelope that we couldn’t open until the exam was over, break included). Given reports I’ve heard, I suspect at least a single digit percent of test takers failed the test this session simply for breaking this rule.
From the test takers who got kicked out of the room and tried to negotiate (unsuccessfully) with the proctors, it was instantly obvious who came from cultures where the consequences of rules are carried out and who didn’t.
> Most western countries have somehow decided over the last couple decades that small negative actions should mostly be free of negative consequences.
There's a general loss of decorum, and it has such immense negative impact. There's so often someone acting like an animal on public transit, which is why many avoid it entirely.
just curious - if they went through the process of providing sealable bags and (I assume) verifying the bags were in fact sealed - why not go one step further and require the sealed phones to all be placed in a bucket which could then be taken to another room to ensure no access, and also no interruptions during the exam from a rogue ring or alarm?
And who's liable if I pay a testing center to properly administer an important test that will heavily impact my future, and during such test a phone rings, another phone's alarm goes off, and another phone gets a series of audio notifications? - when the center could have simply taken all the phones to a another location until the conclusion of the test.
> Existing for zero time would imply it never existed.
In the mathematics of infinity, something can have exactly zero probability yet still happen, and exactly one probability yet nonetheless fail to happen (hence the standard term “almost surely”).
If something can have literally zero probability of existing yet still exist, why can’t it exist for literally zero time yet still exist?
His statement “prior to today never heard much about [Steve Lemay]” leads me to think he doesn’t have intimate access to anyone deeply familiar with design decisions, because anyone who’s spent a little bit of time behind closed doors in that space absolutely knows who Lemay is.
But then he quotes sources who are supposedly “in a position to know the choices”, which would imply they are quite embedded in the design org…
Maybe it’s all voluntary misdirection on his behalf.
If you think about Gruber as a gossip columnist, then it's quite natural that people who just do their job well don't generate much gossip that would reach his ears.
Fair enough, I guess I was just surprised to read from someone who makes their living from Apple insider baseball and Cupertinology that they have “never heard much” about one of the most influential software designers at the company over the last 20 years.
Me feeling is that his sources are mostly programmers and perhaps some managers.
It's also possible he has very few sources left: he's an outsider to the company, and it's hard to maintain sources since people leave, move to different positions etc.
His Something rotten in the state of Cupertino piece earlier this year broke his relationship with the upper echelons of Apple, and to be frank he's no longer inevitable as an Apple commenter. He cannot understand the European Union at all, so he devolves into saying they're idiots. Really? The EU, the whole thing, idiots? Really?
I don’t think we should conclude that a tech columnist who regularly shits on EU tech policy really intends to attack the EU writ large. If he sounds like he’s generalising to the “whole thing,” I’m pretty sure he understands most readers know he’s not talking about the entire EU experiment. Just the backward tech regulation.
Oh no, John criticizes at large. He complains that the EU itself is too obtuse, that the way they enact laws is obtuse, that the nomination process for officials is obtuse.
No. He criticizes quite specific things that show how completely clueless he is, as it is immediately pointed out both by people in the comments and his blogger/influencer/programmer friends.
He really is unique in how consistently wrong and clueless he keeps on being.
Oh, there also was the mildly sensational "I wonder too, what taste Cheetos-dusted 78-year-old testicles leave in one’s mouth. Whatever the flavor, I hope it lingers."
Well excommunication is not appropriate as a term. Apple has cut off access to the most important cardinals for interviews, not told John he's out of the Catholic Church.
> they've lost the wow factor in their software design
Software craftsmanship at large scale is dead, so we shouldn’t expect to see that make a return any time soon.
The last few decades of free market experimentation and evolution have revealed the playbook to maximize engagement+money: sell software as subscriptions, use every means possible (push notifications, full screen ads, etc) to monopolize the user’s attention, prevent users from importing/exporting data to keep them trapped in your walled off app…
In this kind of environment, the little touches and consideration that gave software its “wow” factor are a liability, since everything gets redesigned every 18 months anyway to keep up with the new trends and what A/B testing reveals.
The Apple of the 2000s could offer genuinely delightful experiences because software was in such a different, immature state back then and thoughtful design could be a meaningful differentiator. Similar to how the most successful+profitable games nowadays are filled with loot boxes and dark patterns, and have nothing to do with the masterpieces from a few decades ago.
Indie developers can still make delightful things that treat the customers’ wallet+time+attention with respect (thank God), but those will never make billions and billions the way Fortnite or TikTok or ads in the Settings app can.
> Similar to how the most successful+profitable games nowadays are filled with loot boxes and dark patterns, and have nothing to do with the masterpieces from a few decades ago.
That one actually hurts. I lost touch with games a while ago but it was a good run through the golden era. The cinema is on its way out. At least we have the memories.
Every generation has their stuff, there's new things to be excited about, but the turnover is getting crazy fast.
This feels like the LLM-enabled version of this behavior (except that in the former case, students will quickly realize that what they’re doing is pointless and ask a peer or teacher for help; whereas maybe the LLM is a little too good at hijacking that and making its user feel like things are still on track).
The most important thing to teach is how to build an internal model of what is happening, identify which assumptions in your model are most likely to be faulty/improperly captured by the model, what experiments to carry out to test those assumptions…
In essence, what we call an “engineering mindset” and what good education should strive to teach.
reply