ddj231

> “The essential problem … is what economists call ‘transaction costs,’” Goldstein wrote in an email to The Daily. “Millions of individual copyrighted works may be copied in the course of any training incident, and it’s simply impracticable for the platform to obtain licenses from the copyright owner of every — or, indeed, any — single work.”

Seems strange. The question should be whether you legally have to do so or not. If you do whether it is impractical or not is not a reason to break the law. Perhaps it makes your business not feasible but so be it. If you ran a clothing store that required you steal from other brands and remix it for your customers, you couldn't say "it's impractical for me to get permission", no?

the assumption is that training llms is greater than training individuals. llms may lead to their own "breakthroughs" but so have many people without the use of llms. is the thinking that we'd be dependent on llms and so we are now forced to be only able to discover what llms find discoverable? sounds like a loss

It doesn’t fly though. Not to reduce the engineering feat, but it’s more a jumping car than a flying car.

NYC seems like a bad example since automatic doors haven’t been widely adopted.

you make a good point. on some level these technological advances are about tradeoffs. on the one hand it would be much better for a skilled teacher to take you through a plethora of problems--but since a teacher's time is limited one could make an argument for ai tools.

Though we've had great educations for many years before AI. The cost/benefit is still up in the air for ai use. I for one would not want to spend my time learning from the bot. I find the way it puts words together tiresome. To rely on it for my education, not withstanding its addicting affirmations etc. is not a clear win for me.

Both books you mentioned exist for the curious student. And a plethora of other books. With a good text book you aren't left on your own to figure these things out, the chapters explain the process you need to take to arrive at a correct solution. Combine that with working with your peers, office hours, educational videos, and so much more, and there is a lot for the student. (I mean these bots are trained on an already existing corpus of information)

In the case you described you aren't using AI as a shortcut to an answer but rather to emulate what a teacher would. If you were to then use AI to answer the questions it brings up for you you would have learned nothing in the process. Its like asking your teacher to solve the problems for you. You have an answer but also no growth.

But yes one could argue that using AI not as a means for answering a class's problems but rather to generate more exercises could be valid. Though I doubt students have a lack of practice problems these days.

I find AI is a bad communicator. So it is quite awkward to take advice on prose from a system that generates bad prose, though that could be possible. Still it is unclear, one would have to judge what benefits arise.

Now my earlier point was not that AI is not useful in any way. It is useful with code. Though there is the larger conversation on developer enjoyment etc in the age of AI coding--where the speed of AI generated code is expected, and developers are more alienated from aspects that made coding enjoyable.

Some teachers ban calculator use in their tests, some explicitly check if your work is derived from answers you did not come up with (something you found on the internet etc.). Unless a class takes account of some tool, using it could be to the detriment of the thing you are there for—namely an education.

I’m also for banning it in the sphere of education. First off for much of education the student is not on a quest to discover novel things. But rather to learn known concepts. Meaning the answer is not the point. The exploration of the student is the point. Having shortcut machines to the answer are detrimental.

Secondly it’s unclear whether shortcuts to many “answers” are wholly beneficial in many cases outside the academic sphere. Hence scandals when for example video educators use AI in their material. Or the general push back on AI writing, website design, etc. These answers are being pushed back on.

I think regardless there is a story that goes with the top level of a product. Hence why we have shows dedicated to these things, documentaries, movies, even shark tank. Products have journeys and stories, and part of what is passed from one person to the other is the story--not just the final product. (Not that this has to be a top level thing. and not undermining the labor of many people who are just there to do their job.)

Why I bring this up is because a lot of LLM usage is geared towards building new software products. We one shot some product, and there is no journey, larger collaborated effort. And one may have the feeling "why doesn't this feel as good as building it without this tool".

Yeah I've been thinking that products are a kind of communication. Part of what is communicated is the work, the effort or the journey that it took for the product to come to be. Like when a food product puts the origin story on the back.

LLM generated software seems to push every story to "I wrote some prompts and Claude spit this out". Perhaps that story if fatiguing repeated over and over again. And we're losing something.

On the surface that seems like the case, but after going through a situation similar to what GP described I'm not sure that's the case.

It's like everyone is fighting for your attention. One argues that an educational video is what you should watch. Another argues that building some piece of software you could easily pay a $1-$10 but now custom built for you is a better use of your time.

Also there's the lottery aspect. Where you use Claude thinking that something you build may strike gold. So it becomes like one big addicting lottery ticket for builders.

I wonder if the next addicting tech will be AI mania like you described. Rather than folks watching videos picked by the algorithm as a kind of addiction, folks will be "building" all day addicted to making things with AI. In the AI silo you're a super genius as it eggs you on to keep building

Quitting it completely would solve. The problem is it’s trying to be both a social site and a job utility. You may want the job utility and not the social feed. But they aren’t so disconnected. You land on the site, you’re on the feed. It’s engaging enough to catch your attention, but sloppy enough for you to not like it.

Seems like part of the danger of AI is the ability for folks to put blame for crimes on the AI rather than themselves and thus commit those crimes freely. “I didn’t hack your systems, it was my ai”

I'm surprised at the amount of people here confusing carriers with car seats. It seemed obvious enough to me in the article. It's about wearing babies on the plane, not babies sitting in their own seat.

as addicting as short form content may be its really lacking when compared to fully fleshed out stories. I think big picture there's no way it wins. it'd be like mcdonalds replacing sit down restaurants.

Yeah its also interesting seeing the niche apps that pop up that basically pick what to watch for you and where everyone watches the same thing at the same time. Seems like converging on old school television.

No it shouldn’t. Low risk behaviors should equal low risk outcomes. If it’s safer on average yet causes low risk behaviors to suddenly become unpredictably dangerous, folks will very reasonably push back

the word “decades” is doing a lot of work here. Knowing that Claude is trained on human text is really the load bearing point. I can’t unsee this argument.

Jokes aside, of course it’s trained on human text. The issue is that it has a specific tone which leads to formulaic sentence construction that is tiring. But the plus is that it’s easy to spot if you want to avoid AI generated text.

The second sentence is an ai generation tell. Idk why ai always tries these sort of “wow you” sentences. The tone is pretty off putting once you’ve experienced it enough times.

a bit off topic but the construction of the phrase: some word + "doing a lot of work" sounds so much like a claudism

seems like aesthetically it matches its surroundings. it doesn't seem like an outlier in the neighborhood which kind of defeats the premise of the article.

the issue with some of these quirks is that not that they are not part of the English language but rather the llms usage is unnatural. Also when you have that redundancy proliferated to millions of users it gets tiresome fast, and becomes a tell-tale sign for AI generation. It’s like an AI signature on some written content. Because many people do not want to consume AI generated content this becomes a sign of what to avoid.

Granted picking apart single words outside of context doesn’t give the full picture. (well except “load bearing” which is turning into a meme at this point)

> what's fundamentally different between software running on a machine compared to what's happening in our brains? In both cases you have energy flow following a pattern.

Assuming we’ve scratched the surface of the complexity of the brain. I’d say in one case a human with a will is steering that flow of energy. In the other case it is a probabilistic algorithm steering the flow of energy. The AI is not interacting with world with its own will. I see that as a big difference.

There are presuppositions that go beyond the realm of software engineering that guide one’s views of these things. One is whether you believe the material world is all that is, and that human consciousness is a product of the brain—or that there is such a thing as the soul or spirit of man. From the material perspective you may posit that if you emulate the brain then a sort of AI consciousness could arise. Or that emulating the patterns of the brain equates to emulating personhood. (Though what is material consciousness? I’d say consciousness is by nature immaterial.) I’m not a materialist, and I don’t believe the conclusions that arise from it’s perspectives are accurate.

3 positives from a degree are:

1. the complete demystification of computing. If you’re writing c or assembly and working on OS concepts there’s a sort of aha moment that the computer is not a mystery.

2. The degree pushes you past the hump where a lot of folks get stuck on when learning to program. There’s a seeming spot where things makes sense and you actually understand what’s going on. Forcing yourself to stick with it for 4 years where you’re taught by experts is helpful for that.

3. It gives you a space to explore a wide array of topics. Whether they are useful to an employer or not.

I assume folks can accomplish these 3 on their own. But it would be more or less building your own balanced curriculum and also trusting as you start that your curriculum will work. Whereas within the school system the trust is baked in: if you successfully complete these assignments you will learn foundational concepts.

My view is that these sorts of things are not possible for AI in general. Though we can create things and name them “will” and “desire”.

Software deals with metaphors. Your Amazon shopping cart is a metaphor of a real shopping cart. You your desktop and your file system, etc. are metaphors of real items. But we don’t mistake the metaphor for its object, even from inanimate objects to their software counterparts (shopping cart to Amazon cart).

Now the metaphors are dealing with humanness, things like intelligence etc. And rather than seeing it as software doing what it always does, taking things and creating software metaphors of them, we are starting to say these are actually what they are named. Saying the artificial intelligence is actually an intelligence.

We’d either have to reduce the definition of intelligence such that calculators are intelligent. Or admit that these tools are not intelligent and are rather ways of exploring the work of actual intelligent beings, work that is found in their training data.

I see this line of reasoning quite a bit and it’s a strange one to me. The arguer reduces the sheer complexity of human intelligence and language by saying “we are just running statistical models in our brains” and by doing so makes the leap that Llms are intelligent. It’s an incredible simplification of the human person, who has a deep inner life, a soul, desires, and a will.

I don’t think the aversion to llms as intelligent has to do with the economics of paying intelligent agents more. I’d argue that it’s more fundamental than that. Humans are incredibly complex, and the world of sharing invisible things called knowledge, and the intelligent persons consuming such things which has been going on for thousands of years is far more rich than these synthetic outputs.

When it comes down to it the ai has no inner life, its is dead. A useful coding tool sure. But I wouldn’t call it intelligent.

One side example is just how bad these llms are at artistry. Just saying whatever should statically come next is not good art—and the outputs show it.

Not meaningless because part of the discussion is the issue of anthropomorphizing this tech. When we use language like “intelligent” it carries hints of personhood. People begin sadly treating these things as persons.

We can reap the benefits while clearly telling the consumer this is just a language algorithm.

By that logic you’d have to call other algorithms intelligent.

With more basic algorithms we know that it’s clearly the human programmer and the interpreter of the outputs that are intelligent and not the algorithm itself. For some reason with AI that goes out the window. I believe it should not.

seems OP is not talking about compensation of the deal itself but rather compensation for the actual flight rather than a $100 voucher. Assuming his flight to Brazil was much more than that