[sph]

I am another AI pessimist. Can I please ask the optimists to list the good things LLMs can do for humanity as a whole?

I don't mean banal stuff like Copilot, which is a double-edged sword that might be used against junior developers. I mean world-changing benefits, one step closer to the techno-utopia.

Because on paper, the net benefits vs net negatives, for me and other AI pessimists like the author, are not worth the amount of spam, customer service bollocks and lost jobs LLM will cause to basically make mega-corporations richer.

So please tell me, what will LLMs ever do for us?

[lijok]

Accelerate education at unprecedented levels (if we can figure out how to integrate LLMs into the education system).

To this day I find myself doing a double take every time I'm about to ask ChatGPT to produce a ludicrous amount of output, because it would simply not be reasonable to ask that of some random person. But with LLMs you can do that, over and over again and it will comply every time.

How great is it that if I want to learn about, say, hexagonal architecture pattern, I can ask this thing to produce 500 LOC in a language I happen to be familiar with, and then interrogate it to no end until it clicks for me?

[captainbland]

On "accelerate education" - it's often a bad teacher especially on higher level subjects . I've asked it to summarise research for me using the Bing GPT4 model and it would frequently come to conclusions that couldn't be corroborated in the source material, that were even contradictory across different chat sessions and even generate citation links that were totally irrelevant and incorrect, and then try to tell me that it was of a totally different subject to what it had actually pointed to.

Regular chatGPT is even more dangerous because you have no idea what it's referencing most of the time. Yet it lowers the bar to this poor information to such a degree that people will be incentivised to use it regardless.

[lijok]

That has not been my experience with ChatGPT 4. If I were to ask it about some specialized info about quantum field theory or superconductor tech, I'm sure it would send back nonsense fairly frequently. But if I ask it to explain the difference between median, mean and average, or ask for examples on an architectural pattern, it's seldom incorrect.

[captainbland]

If you prompt directly about the subject it can be fine but it veers into BS territory more often if you try to get it to synthesise information meaning it's less good with applied examples that are given to it by the user. Not always but the error potential is definitely higher.

The problem then I think is that this means that students will have to prompt "on rails" to get reliable answers. The most curious students who want to stretch their knowledge are likely to cause it to generate falsehoods which are presented confidently and convincingly.

[spion]

I'm not sure the error potential is much higher than with average teachers, but I guess it depends on the domain. For coding the error rate is lower than a typical teacher right now.

I suspect we'll soon figure out an effective way for LLMs to look-up reliable references to confirm their answers, which should improve the situation drastically. As they are now most LLMs are very barebones. An educational LLM could e.g. connect to the university textbook library and use that to verify its answers.

[captainbland]

I think the main issue today is its inability to just say "I don't know" when appropriate. A teacher can do that and sometimes that's a better answer than a fabrication. A teacher can also use less certain language when their confidence level is low as appropriate.

I think this is probably a technically solvable problem but doing so has potential 'optics'/marketing issues in terms of reducing the level of confidence it projects which a lot of people associate with competence.

[eletrotupi]

>Accelerate education at unprecedented levels (if we can figure out how to integrate LLMs into the education system).

More likely we'll replace poor people's teachers with ChatGPT and whoever has more cash affords actual teachers. This is so real that we're experiencing this >today<, in a different scale with private schools and distance learning in countries such as Brasil where there are projects to reduce schools and/or move some of them to use distance learning.

[spion]

For most people today, learning / teachers can be far, far worse that even what ChatGPT 3.5 can provide.

[layer8]

I don’t know, but anyone who has the motivation to “interrogate it to no end until it clicks for [them]” already didn’t have trouble to become well-educated before. AI can speed things up for those people, but I can imagine much less impact on general education.

[spion]

The speed difference is out of this world. I have very long sessions with ChatGPT on the Rust ownership and borrowing system. Its sometimes wrong (would say about 1 in 30 messages), but when it is it becomes clear quite quickly and then I search the internet armed with the new knowledge of terminology and relations within that terminology

[lijok]

I agree on your first statement, and raise a challenge on the second.

We're smart, lets figure out how to create the same impact on general education.

Could we wire up ChatGPT so it asks us questions and leads the conversation instead of us? Could we also make it present the questions in ways that make educational youtube channels like Kurzgesagt so easy to engage with?

We could start with private tutoring. LLMs tailored to educate kids of rich parents. Easy market fit as they're already looking for private tutoring. We prove out the need then quickly move onto schools.

Imagine if we could show that engaging with an LLM for an hour produces as much learning as sitting in class for 4. Could we cut school days in half, set LLM interaction as homework for an hour (teachers would get reports from the LLM on how the child is doing)?

[ac2u]

>already didn’t have trouble to become well-educated before

I agree that I didn't have trouble interrogating a subject until it clicked before if I was sufficiently motivated. However ChatGPT reduced the time needed a lot, and thus motivated me to take on more learning than otherwise.

Of course, I take "facts" it outputs with a pinch of salt until I double check things with my own research, but it's great at uncovering unknown unknowns and is like an infinitely patient tutor that will let me throw analogies at it to test if my understanding is correct or needs adjustment.

[jiggawatts]

I just watched the latest Star Trek show and something occurred to me: the Enterprise computer acts much like an LLM. It doesn't take actions on its own, but you ask it any question about the ship, and it answers politely.

LLMs are literally science fiction come to life, something out of the movies.

It's early days still, they're a bit primitive and the compute substrate is woefully underpowered for what we would ideally like... but the promise is there.

[moffkalast]

Working with LLMs while programming is also not unlike the TNG holodeck interface. Verbal generation of the first draft, some feedback and more specific commands to get closer to the intended goal, then sometimes handmade edits to get it all the way there.

[RecycledEle]

The good AI can do for humanity is increasing our productivity.

It should be the goal of every human to eliminate as many jobs as possible.

Every job that is eliminated frees up that person and the people who would have done that job in the future to do other work.

Humanity gain the productivity of that other work.

If we had not forced painful job losses on people, we would still be 97% subsistence farmers, as we were at the time of the American Revolution in the 1770's.

If you oppose progress due to the job losses, please at least be consistent and become a subsistence farmer.

[JeremyNT]

> It should be the goal of every human to eliminate as many jobs as possible.

I mean this is exactly what Drew is disputing.

The capital class will remove the jobs, capture the value of whatever labor is saved, and the workers who lose their jobs will be left with fewer resources and no realistic path to replace their lost income. They won't glide towards some utopian vision where everybody ends up working one day a week on their passion projects.

In some kind of abstract way these ML techniques provide potentially useful tools, but workers will not be the ones to see the benefits of more "productivity" that these tools enable.

The US can't even agree that, despite its vast wealth, health care is something everybody should receive regardless of employment. This country lacks the imagination to handle this situation in a way that improves lives for workers.

[RecycledEle]

If the corporations will not hire you, start your own business.

[falloutx]

Not every worker is so good that they will just get free from the job and they would instantly turn into mega entrepreneur god creating 12 companies in a month. Most of them will struggle to put food on the table, resorting to crimes and eventually ending up in worse jobs

[carlosjobim]

Most of us have had to adapt, with nothing guaranteed. Look at the population under 50, very few in that group have ever had any kind of job security, except for the highly paid specialists. That's been the reality for decades. Thousands of people have lost their jobs because of the hackers on here making effective software solutions. Where was the solidarity then? It's only now when the hackers themselves are threatened that we suddenly have to think about the poor human.

With that said, we should do all to eliminate jobs, but not eliminate workers. Technology can also be used to produce more services for more people, not just producing the same more effectively.

[saulpw]

Subsistence farming is more difficult physical labor, but at least it's more meaningful than being a spam artist or an adtech data engineer or a health insurance denier or any number of soul-sucking bullshit jobs.

[HWR_14]

When was the widespread unemployment in the US due to technological innovations

[incrudible]

The problem is not unemployment, it is raising the bar for valuable labor. Arguably the standard of living has already declined for the average human individual. A factory job used to afford a single worker with modest education a house and a family. That said, generative AI will have far less impact here than, say, a robot arm.

[HWR_14]

That seems like a different (and real, but modern) problem compared to "we never would have left subsistence farming without a massive unemployment period (painful job loss) brought about by new technology" the the OP suggested.

[RecycledEle]

The bar for viable labor is rising.

It is also easier that ever to educate yourself.

These two facts balance.

[incrudible]

If you believe that anyone can learn anything, that might sound plausible to you. The evidence does not point in this direction, however.

[RecycledEle]

Widespread unemployment occurred in the 19th and early 20th centuries as Americans left farms to move to cities.

They were desperate for work.

The unemployment in farms was caused by the Industrial Revolution.

[dandellion]

I'm not a native English speaker and ChatGPT is the best existing tool to spell check and fix gramatical errors in emails. Copilot is another very useful application, and I think your dismissal of it is invalid. Even if we accept that Copilot hurts new devs (which I don't know if you have any proof but I'll make the concession that it is a possibility). A double edged sword hurts and helps the same person. I and half the programmers I know who pay a subscription for Copilot already know how to program perfectly well.

[ChatGTP]

As someone on the receiving end of these types of LLM enhanced emails, I'd say just make sure you actually can be confident it is indeed communicating your message properly.

I work with a guy who has been covering up his poor use of the English language and it's been quite weird to be honest.

Fluent soudning emails full of convincing sounding gibberish is what I've been receiving. Unfortunately the guy just isn't good a communicating his ideas using written language and the LLM can't really fix that.

[dandellion]

When I put in the prompt to correct spelling and grammar it rarely changes anything else. Sometimes I'll use it to reword things but I'll make sure the meaning is clear. I read a lot more English than I write so usually telling if something "sounds natural" is very easy, even if writing doesn't come as easy.

[agentgumshoe]

Does it help you learn better spelling and grammar, or do you just stop thinking about it? Is that a good thing long term?

Same with dev. Does it actually make you better, or does it make you not think about learning so much?

[raincole]

> Does it help you learn better spelling and grammar...

Yes. (Of course I still have a long way to go)

> Same with dev. Does it actually make you better...

Also yes. In terms of "scripting" it helped me tremendously. There were many little tasks that I wouldn't bother to automate without ChatGPT, because the time spent on looking up those niche APIs would outweight the time saved.

Without ChatGPT I would have learned 0 about them. With ChatGPT at least I learned a bit.

[agentgumshoe]

Did you though? Can you remember any of those solutions that ChatGPT provided or why? Or are you now going to ChatGPT more and more because easier?

[raincole]

For English writing I remember a lot of words and phrases that ChatGPT taught me.

For scripting, of course I don't remember the details it provided. I don't remember the details I wrote yesterday either. But at least I roughly remember things like "There is a simple way to visualize any f(x) with about 10~20 lines of numpy and matplotlib. It's not a daunting task at all". I believe these pieces of information improve my decision making a little bit.

I don't think ChatGPT improved the deeper programming skill of mine, like data struct/algorithm/architecture. But I see no reason why it would make me worse in this aspect either.

[danielbln]

Not OP, but for me the higher level decisions stick, but not the bitty gritty detail. And I'm ok with that, I care less about the implementation detail and more about the higher level problem solving, and using these LLMs I feel make me better at that.

[ExtremisAndy]

This is indeed a valid question, but I think it’s largely up to the user. For me, as a hobbyist programmer that sometimes writes code at work to automate certain tasks, I use ChatGPT to quickly create boilerplate/template type code AND to learn how to do new things. When I’m asking how to do something new, I try to actually learn what’s going on so that I won’t have to keep asking about that particular issue. But yeah, the temptation to just say “thanks, ChatGPT” and move on without learning anything is certainly there and could be quite harmful to one’s overall coding skills.

[agentgumshoe]

I think the interesting part of that is, is there money in it? I could see it being useful for hobbyists that don't care about programming that much, but would you pay for that?

[giancarlostoro]

I am not the OP you are referring to, but I am non-native and this is anecdotal but when I grew up and used the internet I learned how to spell way better because of the spell check in Firefox. I really doubt I would have been able to spell half as many words as I can correctly spell any other way. I have some random words still that I infrequently use that I always typo, but theres so many other words that I just know how to write out from muscle memory at this point. My only problem is I dont know how to pronounce all words, I sometimes find a word and get really confused, not too often, but often enough.

I think AI could help me if ChatGPT or whatever else had proper text to speech with a reasonably convincing accent, even if it sounds off, as long as the accents American I am good to go.

I am also a pessimist with AI but I still try it out and use it because I do see it being used more and more.

[spion]

I'm a cautious optimist. If we manage to improve correctness further and set up some guardrails to avoid convincing it to do the work for you, AI can be an incredible augmentation for students and learners, completely transforming education to the level of everyone having something close to a personal tutor

I don't worry too much about the lack of correctness, as long as the learner is aware of the possibility - lots of teachers are sometimes wrong too, and a lot of information on the internet is incorrect.

But there indeed is a certain threshold of correctness. Below that threshold, the tehcnology is more harmful than helpful.

The threshold is also different for different aspects of education - for example for many uses in software engineering education current LLMs are already good enough, but they are nowhere near where they would need to be for e.g. law.

[spion]

I have no idea what the thought process is for downvoting a post which is answering to someone asking optimists to contribute their thoughts.

[ben_w]

> I don't mean banal stuff like Copilot, which is a double-edged sword that might be used against junior developers. I mean world-changing benefits, one step closer to the techno-utopia.

Every (tech) sword is triple-edged, and it's the third one that people trip up on because they never expect it.

That code generation is "banal" to you, or a threat to junior developers, suggests to me that it's already huge.

Likewise the job replacements you fear: that's only even possible when the AI doesn't suck massively.

The current state of downloadable models implies any corporate advantage is temporary.

[Jack000]

A doctor for every person, a teacher for every child, available any time and for free.

[evrimoztamur]

A doctor is a lot more than just a black box taking the patients' descriptions and measurements, and running regressions on them. Doctors can touch, feel, understand, comfort in ways that our sensors or tensors (hah) can't.

Same applies for a teacher too, in various other aspects. Reducing important professions into statistical models is exactly the kind of crappification that the author's talking about. The logical conclusion of perfect sensors and tensors is not here, and the lacking substitutes along the way will be profit-driven, not solution-driven.

[dagw]

Doctors can touch, feel, understand...

Sure they can, but many don't either because of lack of time, or quite frankly, because many doctors are bad at their job. And even in the best case scenario we will never be able to provide doctors to 100% of the population. For many people the choice won't be AI or a (free) caring, passionate doctor who has time to understand you and answer your questions, it's AI or nothing.

Same with teaching. A lot of people simply don't have access to teachers, and if even the ones that do, might not have teachers that have the time and knowledge to actually teach what they want to learn.

[krapp]

This is an argument in favor of more human doctors and teachers, not replacing doctors and teachers with software.

[dagw]

The richest countries in the world cannot even produce enough competent doctors and teachers to fill their current needs. A world that produces enough skilled human doctors to meet everyones needs is even more science fiction than a world with skilled AI doctors.

[tharne]

> The richest countries in the world cannot even produce enough competent doctors and teachers to fill their current needs.

They can easily produce enough doctors, they just don't. A couple of reasons for this: schools inflate the amount of education required so they they can make more money and doctors go along with it (and a crazy amount of licensing requirements) to prop wages up by keeping the supply of doctors artificially low.

You could be an ICU nurse with 20+ years of experiences. Want to make a jumpt to becoming a doctor? You have to start ALL the way from the beginning of med school as if you an 23 year-old humanities major who decided to go to med school. Your 2 decades of hands-on medical experience counts for exactly nothing in the eyes of medical schools and certification boards. Does anyone really believe this is a good way to run things?

[zimpenfish]

Speaking from one of the formerly rich countries (UK), we treat our doctors and teachers incredibly shabbily - long hours, low pay, terrible conditions. It's frankly a miracle that anyone over the last 15 years has gone into either profession.

Fix the low pay and terrible conditions and yeah, you'll easily produce enough doctors and teachers, but late-stage capitalism isn't going to do that...

[krapp]

The richest countries in the world choose not to produce enough competent doctors because of capitalist incentives, not because it's actually impossible.

[ben_w]

By itself it's an argument for both; the argument for "we can't have more doctors" is "we want some of those people to do other things besides doctoring".

[redhale]

The fallacy you are falling victim to, which is common in these debates, is comparing an LLM teacher to a human teacher as a 1-1 replacement, when really you need to be comparing an LLM teacher to what a child has today outside of access to a human teacher: static books and today's internet + search engines.

It's very easy for me to see how an "LLM teacher" developed and trained specifically for that purpose could be of HUGE value over that status quo. That doesn't mean that the child's human teacher goes away, only that they now have access to a new amazing tool at home as well.

[infecto]

Unfortunately most doctors don’t have the time to be that hands on and at the end of the day are just taking your symptoms and comparing it to their flesh database of illness.

There is a lot of value from just having help diagnose/triage people with illness. Certainly not a replacement but definitely a complement to get access to healthcare to more.

[ben_w]

> The logical conclusion of perfect sensors and tensors is not here, and the lacking substitutes along the way will be profit-driven, not solution-driven.

Quite possibly both; governments only switched to universal education instead of having 12 year olds in factories because it was good for the economy, even if some of the lessons are supposed to be good for the (for lack of a better term) "soul".

[Jack000]

I didn't say they would be better. Most people on earth lack any access to healthcare at all. https://shorturl.at/joA23

[RecycledEle]

AIs have been shown to have better empathy then human physicians.

AIs have more patience than human teachers.

[squarefoot]

> A doctor for every person, a teacher for every child, available any time and for free.

For free meaning that it is paid by quietly slipping ads into prescriptions or lessons?

[ben_w]

If the OpenAI pricing is indicative, reading 16k token of medical history and giving a 4k token response will cost $1.50 on GPT-4 and 6.4¢ on GPT-3.5-turbo.

The lower of those two is roughly what someone at the UN abject poverty threshold will spend in 48 minutes 30 seconds on "not literally starving to death".

[tim333]

>spend in 48 minutes 30 seconds on "not literally starving to death".

I don't get that. In the poorest people get by on the equivalent of about $1 per day and not many starve. In fact the only part of the planet where the population is booming at the moment is sub Saharan Africa.

Chat GTP and similar will presumably have free tiers.

[danielbln]

And if you use fine tuning and RAG you can cut the cost by an order of magnitude. Also, how much did it cost 2 years ago?

[ben_w]

Was it[0] available at any price 2 years ago?

(Assuming you mean literally an order of magnitude: 0.64¢ is, judging by Amazon.com, less than the bulk price of a single sheet of unused printer paper, or two thirds of a paperclip).

[0] 3 or anything equivalent to it, given 4 obviously wasn't

[danielbln]

I was making the point that this is new tech, not available to us at all a few mere years ago, so assuming a constant cost when making predictions is difficult. Assume inference prices will go down, not up.

[myrmidon]

I understood the "for free" as "with very low marginal cost"-- and no matter how you socialize healthcare/education, that's not something that humans can match.

[phero_cnstrcts]

Well, if that doctor is fully programmable by the big companies it could just diagnose more diseases and write more prescriptions.

[carlosjobim]

Why the "if"? Doctors are systematically bribed by gigantic medical corporations to write prescriptions for their addictive and lethal medication. Pain killer addiction (opioid addiction) kills thousands.

[krapp]

>A doctor for every person, a teacher for every child, available any time and for free.

I'm sorry... I'm supposed to trust my healthcare and child's education to a piece of software whose primary feature is its ability to effectively hallucinate and tell convincing lies?

And assuming AI is at all effective, which implies valuable (which implies lucrative,) you expect services built on it to remain free?

That's not how anything works in the real world.

[raincole]

No? It's exactly how everything worked so far.

Live performance (orchestra and operas) were for rich only. Beautiful paintings were for the noble and churches. Porcelain was something needed to be imported from another continent. Tropical fruits were so expensive that people rented them.

Now we have the affordable versions of them for everyone in developed countries, and the middle class in developing ones. Yes, often we just got inferior, machine-made or digital copies, but I personally prefer something inferior than nothing.

[krapp]

You're comparing the value of AI versus a human being with the knowledge and skill necessary to earn a medical degree to the value of hearing Mozart live or seeing the Mona Lisa in person to Youtube and JPEGs, as an argument in favor of AI?

>but I personally prefer something inferior than nothing.

Say that again when your AI physician prescribes you the wrong medication because it hallucinated your medical history.

[raincole]

> You're comparing the value of AI versus a human being with the knowledge and skill necessary to earn a medical degree to the value of hearing Mozart live or seeing the Mona Lisa in person to Youtube and JPEGs, as an argument in favor of AI?

Yes, and I think it's a pretty good analogy.

> Say that again when your AI physician prescribes you the wrong medication because it hallucinated your medical history.

I personally prefer something inferior than nothing. I just said it again.

When your human doctor prescribes the wrong medication, would you reach the conclusion that the world would be better without human doctors?

The fact is simple. Professional diagnosing is such a scarce resource that people buy over-the-counter drugs all the time. It's not AI vs doctors; it's AI vs no doctor.

[krapp]

When a human doctor prescribes the wrong medication, it's a mistake. One doesn't conclude the world would be better without human doctors because human beings are capable of thought, memory, perception, awareness, and when they don't make mistakes - and most don't most of the time - it's the result of training and talent.

Meanwhile, AIs don't possess anything akin to thought, memory, perception or awareness. They simply link text tokens stochastically. When an AI makes a mistake, it's doing exactly what it's designed to do, because AIs have no concept of "reality" or "truth." Tell an AI to prescribe medication, it has no idea what "medication" is, or what a human is. When an AI doesn't make a mistake, it's entirely by coincidence. Yet humans are so hardwired with paredolia and gaslit by years of science fiction that such a simple hat trick leads people to want to trust their entire lives to these things.

>The fact is simple. Professional diagnosing is such a scarce resource that people buy over-the-counter drugs all the time. It's not AI vs doctors; it's AI vs no doctor.

That's not a fact, it's your opinion, and I'm assuming you've got some interest in a startup along these lines or something, because I honestly cannot fathom your rationale otherwise. You're either shockingly naive or else you have a financial stake in putting poor people's lives in the hands of machines that can't even be trusted to count the number of fingers on a human hand.

I have no doubt the future you want is going to happen, and I have no doubt we're all going to regret it. At least I'm old enough that I'll probably be dead before the last real human doctor is put out to pasture.

[ben_w]

> Say that again when your AI physician prescribes you the wrong medication because it hallucinated your medical history.

You say that like humans don't err, either though malice or madness or mistake.

https://www.npr.org/2019/10/02/766403612/doctor-gets-40-year...

https://edition.cnn.com/2021/05/18/us/virginia-former-ob-gyn...

https://www.npr.org/sections/health-shots/2022/03/25/1088902...

https://www.theguardian.com/uk-news/2023/aug/22/lucy-letby-w...

https://en.wikipedia.org/wiki/Harold_Shipman

https://www.nationalgeographic.com/history/article/handwashi...

https://en.wikipedia.org/wiki/Andrew_Wakefield

[dagw]

AI physician prescribes you the wrong medication because it hallucinated your medical history.

The big question is will that happen more or less often than it does with human doctor? Human doctors 'hallucinate' stuff all the time, due to lack of sleep, lack of time, lack of education and/or just not caring enough to pay proper attention to what they are doing.

[krapp]

>Human doctors 'hallucinate' stuff all the time, due to lack of sleep, lack of time, lack of education and/or just not caring enough to pay proper attention to what they are doing.

No, they don't. If that happened anywhere near all the time, we would never have given up alchemy and bloodletting, because there would be no reason to trust medicine at all, and yet it works overwhelmingly well most of the time for most people. Meanwhile, AIs hallucinate by design.

[Jack000]

> what will LLMs ever do for us?

Hallucinations are an engineering problem and can be solved. Compute per dollar is still growing exponentially. Eventually this technology will be widely proliferated and cheap to operate.

[Karellen]

> Hallucinations are an engineering problem and can be solved.

I'd like a little more background on that claim.

As far as I've been able to tell from my understanding of LLMs, everything they create is a hallucination. It's just a case of "text that could plausibly come next based on the patterns of language they were trained on". When an LLM gets stuff correct, that doesn't make it not a hallucination, it's just that enough correct stuff was in the training data that a fair amount of hallucinations will turn out to be correct. Meanwhile, the LLM has no concept of "true" or "false" or "reality" or "fiction".

There's no meta-cognition. It's just "what word probably comes next?" How is that just "an engineering problem [that] can be solved"?

[ben_w]

I agree it's more than a simple engineering challenge, but I do so because it is not entirely clear if even humans avoid this issue, or even if we merely minimise it.

We're full of seemingly weird cognitive biases: Roll a roulette wheel in front of people before asking them the percentage of African counties are in the UN, their answers correlate with the number on the wheel.

Most of us judge logical strengths of arguments by how believable the conclusion is; by repetition; by rhyme; and worse, knowledge of cognitive biases doesn't help as we tend to use that knowledge to dismiss conclusions we don't like rather than to test our own.

[Mawr]

How is that bias weird? It has a straightforward explanation - the visual system has an effect on reasoning. This, as well as other human biases, can be analyzed to understand their underlying causes, and consequently mitigated. LLM output has not discernible pattern to it, you cannot tell at all whether what it's saying is true or not.

[zyuiop]

Are they not an inherent problem with the LLM technology?

[krapp]

That's what happened with the internet, which was supposed to be the new Library of Alexandria, educating the world, liberating the masses from the grip of corporate ownership of data and government surveillance, and enabling free global communication and publishing.

It's almost entirely shit now. Instead of being educated, people are manipulated into bubbles of paranoid delusion and unreality, fed by memes and disinformation. Instead of liberation from corporate ownership, everything is infested with dark patterns, data mining, advertising, DRM and subscriptions. You will own nothing and be happy. Instead of liberation from government, the internet has become a platform for government surveillance, propaganda and psyops. Everyone used to have personal webpages and blogs, now everything is siloed into algorithmically-driven social media silos, gatekeeping content unless it drives addiction, parasociality or clickbait. What little that remains on the internet that's even worth anyone's time is all but impossible to find, and will eventually succumb to the cancer in due time.

LLMs will go the same way, because there is no other way for technology to go. Everything will be corrupted by the capitalist imperative, everything will be debased by the tragedy of the commons, every app, service and cool new thing will claw its way down the lobster bucket of society, across our beaten and scarred backs, to find the bottom common denominator of value and suck the marrow its bones.

But at least I'll be able to run it on a cellphone. Score for progress?

[tharne]

> A doctor for every person, a teacher for every child, available any time and for free.

You clearly don't have a deep understanding of what doctors and teachers do.

[danielbln]

Doctor, therapist, teacher, coach - and with the advent of private, fine-tunable models those can be private, local and in the hand of the people.

[franky47]

> private, local and in the hand of the people.

We've seen time and time again that "the people" prefers centralised, paid, convenient management of complexity.

If the majority of "the people" prefers paying a subscription for watching movies or listening to music, I doubt they'll make the effort to learn how to tune and run private LLMs locally, for medical aspects or otherwise. Not when there will be major companies spending billions on marketing more convenient options.

Guillermo Rauch's essay [1] still rings true: it's hard to forego efficiency (though it works just as well for convenience).

[1] https://rauchg.com/2017/its-hard-to-forego-efficiency

[lewhoo]

But I haven't seen any good argument as to why an AI teacher that is even cognitively equivalent to the real deal (and surpasses a human in everything a machine can) won't just become an intellectual worker. I'm not saying this eliminates the need for education, but it certainly erases its major component and that is to be competitive on the job market.

[namaria]

I don't know where this idea comes from that we can get more from language models then what we put inside. Thinking we can process any amount of data and get a competent surrogate mind out of it borders on magical thinking.

[danielbln]

Who is we? The model creator or the user? Getting out what we put in is kind of par for the course in education, yes?

[namaria]

>Getting out what we put in is kind of par for the course in education, yes?

Yes, you put in a person + knowledge and get out an educated person. Is it reasonable to expect to put in GPU + text and get somehow a competent actor, however narrowly we define competence (maybe if we define actor narrowly enough)?

[danleshem]

I think AI has the potential to level up the plain field.

Speaking from an entrepreneur's POV, AI gives me an unfair advantage with respect to the large whales. It lets us complete projects (specifically, software projects) 10 times faster and achieve things before I had to raise millions to achieve them.

I can't tell if the net negative/positive is + or -, but it's not clear-cut for sure.

[sph]

I... don't believe you.

I'm building a product as an experienced engineer with 17 years in the field, and the only place LLMs would be useful without slowing me down is writing copy, whereas I would pay for a copywriter, which I will probably do.

LLM making you 10 times faster to build a product (of which coding is like 20%) needs better sources and proof, as it's a ludicrous number, and a whale with billions has access to much better tools than ChatGPT (i.e actual paid humans).

If by building product you mean "create a basic landing page in HTML", sure. But a landing page is not a company, nor a product you can sell.

[turtles3]

Does AI really speed up projects by 10x? In my experience, tools like copilot help with the smaller things, but it's bigger things that really matter in a project.

For example, the right database schema or the right architecture I could easily see saving you 10x the development effort, but these are the things copilot is least able to help you with.

Am I misunderstanding something here? Or could it be that what you are finding is that your experience is helping you build faster, and this is being misattributed to AI?

[BoorishBears]

When you say someone should list all the good world changing benefits to counter your suggestions, it means you're expecting world changing negativity.

That means you're envisioning a future where LLMs are so powerful that they actually cause huge societal impacts but somehow they can't detect spam, or make customer service more effective, make the general populace smarter?

You're asking a for an answer that you by definition wouldn't accept. To me LLMs are like cloud computing: cloud computing technically didn't change the world in that your average person knows what a load balance is... but there was a democratization that allowed many great things to be built at a scale that was previously not possible, and the results of that are what brought value.

[aragilar]

Here's one positive use of AI: detection of shark species by drones to inform lifesavers as to whether a beach should be closed or not. This saves both shark and human lives, and is likely more cost effective than other methods (e.g. nets). Noticeably, it's the choice of data (i.e. not ingesting the labour of others without recompense), and how it's used, that make it very different to the various products that startups are producing. But it's the startups (and their owners) that will profit from misuse.

[funerr]

There are too many variables to predict the future. Yet, I believe that it will empower more 1-3 person businesses to compete with the big incumbents.

Imagine you could run teams of marketing/sales/support ai-agents and focus on your core business. "The next 1B, 1 Person companies".

Just 15 years ago you'd need teams of 10+ people to do basic ml tasks you could do today. Now you have an LLM or a SaaS service that outperforms those teams for the fraction of the cost.

[supriyo-biswas]

I expect such gradients to be temporary. After all, there’s nothing that stops the 100,000 person mega corp to do something with an LLM either.

[calebkaiser]

We will certainly see increased adoption at large corporations, but where LLMs will have a lasting equalizing effect is in areas where scaling has a logarithmic sort of benefit--very impactful going from 0-1, but less and less as scale increases.

To give a concrete example, if you're a standard 1-3 person business with a niche enterprise SaaS product targeting the enterprise, outbound sales is likely going to be a meaningful channel for you. Having a team of good SDRs is a huge step function for your org, and if an LLM can give you that functionality without you needing to hire, then they'll make an outsized impact on your business. However, the Google-sized org operating in your space likely already has a huge sales and marketing org. They've likely "saturated" the market in the sense that they're already engaging most of their prospects in one way or another. Adding the functionality of many SDRs via an LLM might give them some efficiency gains, but it won't unlock a new level of growth for them, nor will nullify the value the LLMs provide your smaller org.

[Paradigma11]

Empower artists to make far more complex works that they would never have been able to do alone before.

One artist might be able to produce a complete animated movie controlling all aspects.

[pjerem]

As someone suffering from severe inattentive ADHD, I can see a great potential in the rise of really smart personal assistants which could aggregate every single data point in my life and throw the information at me when needed.

But that’s the only thing I could imagine and it must be local not to be a privacy nightmare. Because I have no doubt that Google and Amazon are already working on that.

For anything else, I just agree with you and the author. It will just make capitalism worse.

[projektfu]

This is kind of what kills me about the current crop of assistants. They all make my ADHD life harder, not easier. The Google Assistant, for example, cannot handle the few obvious things I ask. "Send an email to Jim." Nope. "Record my weight in Google fit." Nope. It won't look at any calendar that's not a Google calendar. It puts other people's flights on my schedule. It tells me I've got good traffic an hour before rush hour, but doesn't give me an update as it gets worse. I've probably wasted more time trying to get it to work than it's ever saved me. And it's twice as good as Siri was.

I'm always getting reminders about bills I've auto-paid, that Google has dug up from the bowels of my email, and yet I still find myself scrambling on things I forgot about.

I'm highly unimpressed by anything other than the incredible confidence with which these systems will lie. Online and phone support is bad enough, AI will not make it better, not because they can't make an AI that could figure out that you need a technician to come, but because corporate won't trust the AI to not deliver a pony along with them. And who could blame them? As soon as they give AI any power it will become the tool of the stainless steel rat. Instead, it will continue to be used to fend off support calls and fudge internal metrics, just like the last round of automated support.

[pjerem]

> The Google Assistant, for example, cannot handle the few obvious things I ask.

Oh my, I think you are not ready for Siri.

[supriyo-biswas]

If the model of streaming services is anything to go by, the LLM would be offered as a paid subscription with ads.

[yomlica8]

And once ads are involved, the perverse incentives will mean the assistant will begin pushing you toward advertised products.