Hacker News Reader

Comment by maxdoop | Hacker News Reader

For a supposed smart group of individuals who are good at leaning into the latest innovations, I feel most HN comments as of late are completely , ignorantly dismissive of LLMs simply because LLMs aren’t “amazingly perfect” in their current state .

Yet… we are are talking about a tool that came out literally FOUR MONTHS AGO. And the huge advancement on that came out TWO WEEKS AGO.

Yet everyone here continues to proclaim, “it’s not even that good , honestly.” As though no progress will ever be made from this current moment in time.

I feel like I’m taking crazy pills.

View on HN

This is honestly so interesting to me because I feel like I have the exact opposite experience. I have never experienced the level of hype there is for LLMs both here on HN and in society at large. I see people talking about how amazing they are, how they will replace everyone in a matter of years and even achieve super intelligence. I see people argue strongly against even the mildest LLM skepticism. I use these tools every day because they are useful but I'm also blown away by how often the answers are completely wrong, like not even close. I sometimes feel like I am crazy or everyone is using an entirely different product to me.

But the fact that you feel the exact opposite shows that maybe this is just an artifact of cognitive bias.

I guess we are two then, I feel the pretty much the same way as you. I might be completely wrong about all this, but it is getting to a point where it feels like I'm getting gaslit when I compare it with my own experience with these tools so far.

Crafting good prompts makes a big difference. Also, after it gives you an answer, say you think there's an error and ask it to check its own result. It will often correct itself. This self-reflection ability leading to more accurate outputs was pointed out in the new Reflexion paper:

https://nanothoughts.substack.com/p/reflecting-on-reflexion

https://twitter.com/Orwelian84/status/1639859947948363777

Here's another good thread about it:

https://twitter.com/ericjang11/status/1639882111338573824

Thanks for the links, appreciate it!

Genuine question as I definitely need more to balance my own thinking —

What has your experience with GPT been? For me, GPT-3 was not really useful as a software dev.

But GPT-4 is miles ahead of that. It’s helped me write code maybe 4-8x faster than usual, and has even allowed me to debug existing issues far, far quicker and more accurately than I’d ever be able to on my own.

Part of the gap very well might be my own mediocrity with development . I wouldn’t argue that folks with far superior skills and novel challenges day-to-day might be unimpressed.

But as an average dev writing pretty boring code (REST APIs and system integration mostly), I’ve been blown away by GPT-4. I am pretty well compensated and have been in the field for 10 years, too; but I am aware of my own shortcomings.

I've rarely seen completely wrong answers. When I did see them it was because I was too prescriptive.

Like, I might ask "using this library, implement that feature" in the hopes that there it has learned of some way to do a thing I haven't been able to figure out. In those cases I see it hallucinate, which I assume means it's just combining information from multiple distinct environments.

If I'm not too specific, it does a pretty good job.

IMO its biggest fault is that it is not good at admitting it doesn't know something. If they can crank up the minimum confidence values (or whatever, the values used to guess the next token), maybe we'll see better results.

> But the fact that you feel the exact opposite shows that maybe this is just an artifact of cognitive bias.

Or social media/advertising engagement algorithms doing their work. After all, you get more engagement with negative emotions than positive ones.

It seems plausible that you and the OP are at least slightly on different sides of the LLM issue, and so you and the OP could literally be seeing two different realities crafted by engagement algorithms, because it detects that each of you pay more attention to the other side that you disagree with, and that snowballs into seeing only the other side and thinking you're taking crazy pills.

It depends on which one you use. Bard for example is kind of crap.

> As though no progress will ever be made from this current moment in time.

The issue is that the folks promoting chatGPT are for the most part incredibly dishonest. E.g. this entire blog post is about the AI having written a sales email, with zero written about how well it actually converted. The author is claiming that the AI can do a superhuman amount of work in 30 minutes, but we don't actually know if it did any work at all.

How can we even know whether OpenAI is making progress if we don't know how good it is in its current state? Back when Go AI was far less good than even the average club player, we at least knew what rank the AI was playing at. Whereas right now the ChatGPT equivalent is basically that it's putting stones on the board in a way that looks somewhat like a real game, but you're not allowed to know what level it's playing at.

I don't understand this perspective, the only way it makes sense is if you haven't bothered to really try to use the tool to get something done. There are literally a hundred million+ people out there using it to get shit done. I didn't sign up until the GPT-4 release because I had tried 3.5 and thought it was ok but limited, I had more luck using the API and custom prompts with GPT-3 but GPT-4 converted me. I've been using it daily and for increasingly more stuff. It's so good that I'm actively trying to break the habit of using search engines to look up docs, error messages, and SO answers. Mandatory disclaimer: it doesn't always produce perfect output that works on the first try but you know what? the old search result method doesn't work on the first try either especially if I have to read bad or limited documentation, parse a somewhat related SO answer into the solution I need or search through miles of issues about the error message where everyone is just complaining and no one has found a solution. The big difference is that GPT is a hundred times faster, provides better explanations and even when it's wrong it gives me enough information to figure out the problem on my own.

Sorry but it's just frustrating seeing "how can we know it's better?" when it's right f-ing there in front of you. Maybe you don't want to spend $20USD to try it out, fine whatever, wait until it's free to use but don't make lazy negative remarks from a place of ignorance.

Exactly. It’s similar to someone saying Google can do a superhuman amount of work.

“All I had to do is search for a sample product launch email and POOF it appeared! Just had to fill in the company name.”

It’s the same thing with the minor code snippets being “written” by Chat GPT. Any real programmer knows that Google could give you pretty much the same thing. And they also know how complicated their actual job is that goes well beyond the simple prompts people are using that everyone has been googling for over a decade now.

It’s all hyperbole. This technology is just an evolutionary improvement on Google.

That's one application. You don't find it at least a little jaw dropping that the same program that is "an evolutionary improvement on Google" for programming tasks can also translate text, write (bad but improving) poetry, give reasonable relationship advice, creatively answer questions of the form "what do <x> and <y> have in common" for even radically different x and y, and correctly solve analogy puzzles? All in fluent English?

English? I've been prompting GPT-4 in Spanish all along, and it responds in impeccable Spanish. I can ask it to change the writing style of an essay to famous Latin American writers and it does it admirably. Translating back and forth between English and Spanish respecting meaning and context, even subtleties, no problem. Better than DeepL.

If y'all have been focusing on GPT-4 coding abilities, I ask you to try it with literature-based prompts. GPT-4 is an exceptional writer, summarizer, and style corrector.

I'm not going to claim I know where this is going right now with LLMs, but: throughout history, I don't think anyone really saw automation coming for their own jobs until it happened. I think people tend to think of their craft as an 'art' rather than a product. They miss that other people would MUCH prefer it to be automated, and that people will be willing to put up with slightly inferior output if it's practically free in comparison.

This is already a meme but I'll repeat it here "hard pills to swallow: if you're afraid of ChatGPT taking your job then maybe you're not very good at your job"

What this means to me is that if you see the parts of your job that are basically drudgery as being what you're good at then you should worry. If instead you see them as a necessary evil that you have to do in order to accomplish the bigger goals you're aiming for then you should be happy.

That's the kind of take that I was referring to. Yes, if GPT could replace you now, your job is not very skilled. But if investors see a signal that software might be automatable, they are going to throw mountains of cash at that and they'll be gunning for your job as much as anyone else's.

I've seen a lot of comments on here like "well LLMs are good at writing snippets, but could never write or maintain large codebases to accomplish a larger goal" - but are you SURE about that 'never' given the current generation is already doing things people suspected they could never do? "But they'll never be able to manage / interpret stakeholders", Are you SURE that LLMs will have to adapt to fit stakeholders, and not the other way round? I don't know for sure, and even if this is coming, I've no idea on the timelines. But I'm not completely writing it off as a possibility anymore either.

> I've seen a lot of comments on here like "well LLMs are good at writing snippets, but could never write or maintain large codebases to accomplish a larger goal" - but are you SURE about that 'never' given the current generation is already doing things people suspected they could never do? "But they'll never be able to manage / interpret stakeholders", Are you SURE that LLMs will have to adapt to fit stakeholders, and not the other way round?

That sounds a lot like the self driving cheerleaders five or ten years ago. That work so far has resulted in some awesome features like adaptive cruise control and parking assist but it fell far short of what the hype was promising to deliver by now.

Five or ten years later Mercedes is the only company getting ready to ship level three self driving. Level four and five are still a pipe dream, practically restricted to a few companies like Waymo in a few controlled environments like Phoenix and San Francisco.

GPT4 is great and I can't wait to see what 32K or even 100K/1M token models can do, but I fear we're about to hit the point where progress grinds to a halt because going further requires something closer to AGI than what we have now.

As I say, I'm not sure about anything here, I just think it's risky to write the idea off right now just because it requires progress we can't forsee.

I also don't think the comparison quite works, because no one is saying that we need to get down to zero humans for this to to be profoundly disruptive, just enough humans to code review and make relatively small changes, I wouldn't be amazed if that's what software engineering becomes in the coming decades.

I think there's a good chance that the next version of GPT will be better at writing code than a junior level dev. I don't think this is going to result in a net loss of jobs. As software development becomes less expensive per unit volume I think demand will increase so much that both the total number of jobs and average salaries are going to increase.

Yes, because our entitlement expands even faster than GPT's capabilities.

But it's not free. They keyholders might be making it cheap right now but once the quarterly reports aren't showing enough growth it'll all go up.

I don't buy that at all - you should look at what happened in the past, not to what's happening in the present. Over time, prices will go DOWN not up, due to improvements in hardware, engineering and probably smaller, more specialised models. A lot of investment will go into making that happen, like any other automation. If OpenAI did want to try and price gouge, then they'd just make a gap in the market for someone else to fill. Also - just consider how absurdly high SWE salaries are - it would have to get orders of magnitude more expensive than GPT 3.5 (say) for it not to be 'basically free' in comparison.

People are generally very resistant to change. Our minds are literally programmed against it...acknowledging an enormous change happening very quickly triggers all sorts of yucky things in our lizard brains. So we try and rationalize things and fid some way of explaining to ourselves that "no, this isn't actually a big change, my model of the world is still rock solid" as a pure survival mechanism.

I would have thought filtering people on "being in tech" would have weeded out the people resistant to change, so I'm surprised by the negativity on anything innovative here, whether it's EVs, self driving cars, AI, crypto (although I can forgive this one) or VR.

People in tech are still just people.

100000% this. HN attracts some smart people but they're still often very flawed people. Myself included.

LLMs were invented four months ago?

Honestly, I feel the opposite. I'm sick of the endless fawning over ChatGPT because it can print code that exists in a stackoverflow answer somewhere.

I'm also cautious about extrapolating. Constant improvement, let alone exponential, is far from a guarantee, but from what the LLM acolytes would have you believe it's pretty much a given that GPT6 will be an AGI before 2030.

I feel like I’m taking crazy pills

I feel the same way but on the other side. All I see are non-technical or quasi-technical people using AI tools to perform work that is x% better than Google could do.

The only entity at risk of being displaced from this technology for the foreseeable future is Google.

The only thing left to discern is what % better this technology is than Google’s antiquated algorithms.

Think about it for a second. Putting the right query into Google could give similar templated results. This technology is just an evolutionary improvement on that.

That's just wrong, have you used GPT-4? Last I checked I can't ask Google to invent a new programming language specification for me and get a result in under a minute. I then asked it to write a short tutorial for that language in the style of Learn X in Y minutes and got a good result. I then asked it to write bubble sort in the new language and got a result. Comments like this make me feel like I'm taking crazy pills but really I just know that there are a bunch of lazy curmudgeons out there that can't be bothered to look for themselves.

That's just wrong, have you used GPT-4?

Yes. A lot.

Last I checked I can't ask Google to invent a new programming language specification for me

I have literally never done this in my job. Ever. I've been a professional software developer for multiple decades.

I then asked it to write a short tutorial for that language in the style of Learn X in Y minutes

So you can't Google, "Tutorial How to Learn X in Y minutes", get a result that has a completely viable format and details and then do the remaining 15% of the work to fill in your specific items?

I then asked it to write bubble sort in the new language and got a result

Do you know how many times I've written a bubble sort in my actual software profession?

The use cases you're bringing up is very typical of what people bring up with GPT-4. Contrived stuff without real-world application or something that is an evolutionary improvement on Google.

Overinflated claims of what GPT-4 can do causes real harm to our industry. There are business owners making decisions now based on the smoke and mirror demos that people are showing that has convinced them that they won't need developers in a few years. I literally met with a computer science major who was considering switching majors because of GPT-4.

Over the next decade, GPT-4 will be an amazing productivity enhancer for actual software engineers. Just like Google has been but with significant improvements.

> There are business owners making decisions now based on the smoke and mirror demos that people are showing that has convinced them that they won't need developers in a few years

If they aren't doing their due diligence on this then that's their problem. If my example isn't good enough then provide one of your own that you couldn't get to work.

While I’m aware of my own mediocre programming skills and by no means claim to be an expert, I’ve been a software engineer for 10+ years and have done pretty well for myself. I’d hope that doesn’t qualify for “quasi-technical”, but maybe you are lucky enough to work on novel problems that require advanced problem solving every day.

Regardless, I’d argue that GPT-4 is actually far better at programming assistance, understanding concepts (it’s phenomenal at explaining things when prompted within a context), writing in general, and kick-starting creative pursuits than it is being a Google-replacement (for now, at least).

Have you seen or used GPT-4? What has your experience been? What has it failed at, or rather, what would you wish to see in such a system that might make you to, “huh, ok — that is pretty cool.”

I was just thinking this same thing.

I think the sentiment is warranted in some contexts, but in others it just seems dismissive.

For instance, I am not impressed by ChatGPT's code output. It seems to be incapable of understanding the nuance that is required to modify known or similar solutions to fit a novel problem. In that sense, I don't think it's doing much more than a search engine. It could be it just hasn't had enough training examples. It could also be that there is something uniquely more difficult in regards to solving novel problems via code (I doubt this).

But to get a marketing campaign and a website (albeit in mediocre quality) from text prompts is truly amazing imo. A lot of people are missing the point that these models are in the toddler stage of their life.

Completely agreed. Plenty of criticism focuses on specific algorithms or frameworks, as though there’s something inherent to LLMs that render them unable to improve in that regard.

Other objections, such as “if your job can be replaced by an algorithm, you weren’t particularly valuable in the first place” or “software development is much broader than writing code” are irrelevant to the question of whether a large portion of developers will be replaced.

I don’t think they will, given the world’s appetite for software, but it might become a less prestigious and lucrative profession on average.

Why do you think it will become less prestigious? This technology is going to 10x the abilities of every developer and since we are now all guiding an AI to write code instead of typing it out ourselves it's going to become even more important that software engineers earn the "engineer" part of their title through education or experience. More than ever we will need to know how to describe the problem and the solution and how to think critically about the code that the AI produces. The people with the knowledge to do so are going to become even more valuable since they will be among the most productive but even the fresh out of bootcamp junior will be more productive as a prompt-monkey instead of a code-monkey.

My reasoning starts from the idea that “prompting” is essentially programming in a higher-level language, and it will eventually replace traditional coding. At a certain point, we’ll be stating business requirements, and the algorithm will take care of logic and resource provisioning.

Longer-term, what we now consider tech skills will be replaced with communication skills and business domain knowledge. This will cause an influx of workers from different professions and walks of life. As the field starts encompassing a broader spectrum of work, the barrier for entry will be lowered, and there’ll be more work and more practitioners.

There will still be high-paid jobs, but on average, software development will become a more traditional middle-class profession.

This is all speculation on my part, of course.

People were saying this exact thing in 1960 about COBOL. The job of programmer was going to go extinct because business users could write their own software.

We've been through a number of iterations of the same pipe dream since then, but it always turns out that the actually hard problem in programming is figuring out the requirements in full detail without handwaving and glossing over anything, and translating them into unambiguous instructions. And "workers from different professions and walks of life" just inevitably suck at that.

Whether this time it really is different will hinge on whether LLMs can really figure out the handwavey parts, or whether those will be exactly where they will always make up shit and be confidently wrong.

No, higher-level programming languages didn’t make developers obsolete, but they’ve been increasingly lowering the barrier for entry into the field. Plenty of business people can write basic SQL queries nowadays, and Python is jokingly referred to as pseudocode.

Even before the recent ML advances, there’s been a shift towards involving a broader, less skilled workforce as the amount of work expands. Hence the people switching careers and getting gainful employment after a few months of even weeks in a coding bootcamp. Don’t think that was as common in the assembly coding days.

So while none of those advances suddenly destroyed the profession, there’s been a gradual change to include a broader spectrum of practicioners. I don’t expect the LLMs to revolutionize the field in two months, but I feel like it’s safe to extrapolate that this is where it’s headed eventually.

You’re implicitly trying to make a prediction on where the technology is going based on present trends, something that the industry has never been good at.

Remember how 3D printing was supposed to be ubiquitous by now? Or how self-driving cars would lead to an economic apocalypse because of how reliant the economy is on truckers? Remember all the predictions of bitcoin going to $100k? Remember how an AI startup called “The Grid” was making news for their AI website builder back in 2016 (edit: 2014)?

Don’t even get me started on VR and mixed reality. Remember Magic Leap?

My takeaway from all this, other than that futurists are hucksters, is that progress is actually quite slow, and relies on sudden, unpredictable breakthroughs. I mean, without the iPhone, we arguably wouldn’t have smartphone apps, the death of Flash, or responsive websites. Arguably, without this single pivotal product, tech as we know it would be a very different place. I know it’s a stretch to some to call the iPhone a breakthrough, but it’s impact has been pretty huge.

The catch is knowing when something is a real breakthrough, and when it isn’t. I genuinely thought that the Oculus was at the time, and yet here we are, years later, and it seems like nothing has changed, aside from incremental improvements in VR display technology and a very niche VR gaming community.

GPT-4 is clearly impressive from a technical standpoint, useful even, but where does it really go from here? Does the technology take off, or does it plateau in its present state?

Is it VR and 3D printers all over again?

And remember Transistors?

Paradoxically the hype was much lower for transistors.

The root of the problem is that no matter how hard you want to believe, the LLM doesn’t actually know anything.

If you gave an idiot something very intelligent to say and he read it out loud perfectly, people might be very impressed too. That’s GPT.

I used to think this but now I'm fairly convinced that it "knows" somewhat less than someone who was locked in a tiny dark room with no input except the ability to read a lot of text from the internet would know if that was their whole life. I don't believe it has a sense of self or consciousness, just that it possesses whatever knowledge is embedded in written text. Maybe a better analogy would be if you could cut out someone's language centre and put that into a jar hooked up to text input and output. It's not a whole mind but it sure feels like it's a piece of a mind that can do mind-like stuff.

You can compare GPT-4's limitations to Hellen Keller's. Someone who is deaf and blind can still reason as well as someone with "all inputs enabled". Hellen Keller still had a "full mind."

That's why I included the part about being locked in a small room with only text for input. People who are deaf and blind still interact with the world through their other senses. GPT-4 has no other senses and no body.

That’s just the illusion of LLM drawing you in deeper. Knowing the correct thing to say is not the same as knowing things.

And what is “knowing”?

Everyone repeats the retort you gave, yet I’ve yet to see a clear definition of “knowing”.

Are you talking about the difference between memory and reasoning? It's a bit hard to understand what you mean by knowing the correct thing vs knowing things. Both mean you know things, correct or not.

When you know something you know it in multiple different contexts and forms, not just how it relates in response to a stimulus, or a prompt.

Still not sure what you're talking about that's different from what GPT can do. It's very good at transferring from one context to another while retaining the same intent or meaning. Could you give an example of something you think it can't do?

Innovation.

Could GPT be given some screenshots of a game you want to play and then code it up?

Could you run through a demo of some competitor’s app and have it make something similar but better?

I'm using it right now to help write a game that I had an idea for. I'm writing in a programming language that isn't the same one I use daily and I'm using a graphics library that I've only used once before to make a small game and GPT has been a massive help with this. It's helped me solve some tricky problems like getting a shader to work the way I wanted it to and I've used it to create first drafts of all of the code so far. I guess that's not pure innovation but it sure as hell has a better grasp on a lot of the stuff it's writing then I did at first. It can't just look at a picture and produce the exact game you want but neither could I. I'd have to ask you a bunch of questions about what you wanted the gameplay to be like, if you wanted to release it for PC or console or both, I'd have to get an artist to create a whole bunch of concept art and then ask you to approve the ones you like and then I'd need to implement all the code and play test it with you and make changes etc. It's a bit unfair that you want this tool to do more than a single person could just to prove that it "knows something". Just because it isn't 100% autonomous doesn't mean it has 0 knowledge or ability.

Read the Microsoft Research paper on GPT-4. Some extremely intelligent behavior emerged from just text that has nothing to do with the text it was trained on.

But what does it mean to “know” something?

And when something gives the increasingly-accurate illusion of knowing, I fail to see how it matters (with regard to impact on society and overall utility).

I’m not saying GPT-4 is this amazingly accurate, near perfect model. But if you extend the timeline a bit, it’ll be able to become more and more accurate across a broader range of domains.

Furthermore, how can we prove a human “knows” something?

Ask anyone who has hired someone who says all the right things and seems intelligent, but has no experience or skills in what they actually talk about.

When I write code, I don’t just focus on solving the problem at hand, I think about things like, like: how will another human interpret this, how maintainable will this be, what are the pitfalls down the line, what are the consequences of this, any side effects, performance implications, costs, etc… things GPT does not know.

But still, what does it mean to know something?

And your point about humans lying about knowledge only to be found inexperienced is quite the opposite of an LLM (albeit there is the hallucination problem, but GPT-4 is a massive improvement there):

These models do have “experience” aka their training data. And I would argue with most every one of your examples of things that GPT doesn’t know.

You can ask it about performance implications, side effects, costs. It’s quite good at all that right now even! Imagine the future just a few years out.

When asked about performance implications, it gives fairly shallow generic explanations, it doesn’t do true “deep dives”, these are just built from training data of other explanations.

There is no “getting better” from this. If you gave a monkey a type writer and it occasionally typed words randomly you wouldn’t say “Wow this is just what it can do now, imagine several years out!”

Continue asking it to provide details and it can. Or , prior to asking it about performance, ask it to respond with as much detail as it can and have it include details you specially want to see.

Comparing GPT-4 to a monkey with a typewriter , and claiming the absolute of “there’s no getting better from this” when we’ve literally seen dramatic progress in just months?

I think you’re missing out on some of the utility this stuff can actually provide .

No, you see it needs to do these things on its own, unprompted. It has to consider multiple solutions to problems it encounters and choose the best one, not just the most probable one. It’s not made to evaluate things that way, you can’t hand it multiple implementations and ask it to weigh the pros and cons of the different approaches and recommend the best one for what you’re trying to do. You can’t hand it your code for code review and ask what you could improve and expect to get a response that isn’t just fabricated from what other people have said in code reviews.

And it will never do those things, because it’s an LLM and there are limits to what LLMs can do. There is no “getting better”, it will only sound better.

If it’s going to replace programming, the prompts simply cannot be more laborious than writing the damn code yourself in the first place.

Think 50 LLMs with different personalities and focus points talking to each other, mixed with stuff like Wolfram. You can instruct them to “invoke” tools. An outside system parses their “tool use” and injects results. You can get quite crazy with this.

LLMs are just the part of a much larger looping system that can do these things you speak of. Be active and seek out stuff. Of course, it’s all illusory, but I’m sorry I think it’s no different with myself.

By the way, it actually gives ok reviews on novel code, so I’m not sure what you mean. At some point nothing is truly novel, even innovation is composing existing “patterns” (at whatever abstraction level).

I would like to see examples of these “ok code reviews”. Everything I’ve seen has been fairly plain and not too insightful.

> There is no “getting better” from this. If you gave a monkey a type writer and it occasionally typed words randomly you wouldn’t say “Wow this is just what it can do now, imagine several years out!”

So thinking that chatGTP could gain understanding is as crazy as the idea that primates could learn to use tools or type words?

What does it mean to "know" something, and why does GPT not fit the criteria?

An AI version of Popehat's Law of Goats applies. If an AI attached to a nuclear command center issues an order to launch the nukes, somebody (likely including us) is fucked. It doesn't matter if the AI "knew" anything, if it deliberated over its target selections, if it assessed and prioritized the most likely threats. Something happened and now we have nukes in the air.

If an AI reliably produces good code in response to prompts, it is a better program than most humans. If an AI reliably produces prose that's free of errors, well organized, and summarizes the issues asked for, it is a good writer. Irrespective of what we can say it "knows" or doesn't know.

Yep, but that’s true for HN and almost every new piece of tech. The general reaction is always “meh, it’s not 100% perfect so it should be trashed”.

On the other hand we finally got full self-driving cars by now. Or did we?

https://jalopnik.com/elon-musk-promises-full-self-driving-ne...

If you’re on HN, I would assume that unlike regular customers, the journey towards a technical achievement should be as exciting to you as the final achievement itself.

ChatGPT is a giant step forward in the journey towards AGI. Tesla’s cars, for all their flaws, are big steps forward in EVs and even self-driving.

I personally find that exciting enough.

> Yet everyone here continues to proclaim, “it’s not even that good , honestly.” As though no progress will ever be made from this current moment in time.

As a non-expert in the field I was hesitant last year to disagree with the legions of experts who denounced Blake Lemoine and his claims about Google's AI being alive. I know enough to know, though, of the AI effect <https://en.wikipedia.org/wiki/AI_effect>, a longstanding tradition/bad habit of advances being dismissed by those in the field itself as "not real AI". Anyone, expert or not, in 1950, 1960, or even 1970 who was told that before the turn of the century a computer would defeat the world chess champion would conclude that said feat must have come as part of a breakthrough in AGI. Same if told that by 2015 many people would have in their homes, and carry around in their pockets, devices that can respond to spoken queries on a variety of topics.

To put another way, I was hesitant to be as self-assuredly certain about how to define consciousness, intelligence, and sentience—and what it takes for them to emerge—as the experts who denounced Lemoine. The recent GPT breakthroughs have made me more so.

I found this recent Sabine Hossenfelder video interesting. <https://www.youtube.com/watch?v=cP5zGh2fui0>

I’m with you. I believe we haven’t even got started in exploring what you can do with the existing tools, let alone what new and improved ones we are about to get.

I can imagine LLMs becoming a UI in front of just about everything. Instead of googling the ffmpeg flags you need, you’ll just ask your terminal to walk you through it.

> Yet everyone here continues to proclaim, “it’s not even that good , honestly.”

We must be reading different HN. This is not at all what I'm seeing. As of now, the first comment I'm seeing which is dismissive is the sixth from the top, while your comment is second from the top.

I feel like it's a tale as old as time that people have very different experiences reading HN despite it being the same content for everyone. I assume due to some elements of human psychology.

Ask an HN person if the website is pro-spaceman and anti-spaceman. A lot of people feel very strongly about one way or the other.

I would be careful with the extrapolation... - https://xkcd.com/605/

> who are good at leaning into the latest innovations,

maybe not true.

There are many threads where new products got comments of "useless" and then launched. HN users are smart, picky, and not representative for crowds.

This might be how the singularity plays out.

I asked Google Bard to tell me a short story about a computer hacker. Here was the result, completely unedited... Blew me away:

Sure, here is a short story about a computer hacker:

Once upon a time, there was a young hacker named Kevin who loved to explore the world of computers. He was always curious about how things worked, and he loved to find new ways to break into systems. One day, Kevin was exploring a new website when he came across a hidden file. The file contained a list of passwords, and Kevin was curious to see what they were for. He tried a few of the passwords, and to his surprise, they all worked! Kevin had just hacked into a major company's computer system.

Kevin was excited about his discovery, but he also knew that he had done something wrong. He had hacked into a system without permission, and he could get in a lot of trouble. But Kevin couldn't help himself. He was too curious about what he could find in the system. He started exploring the files, and he quickly found out that the company was up to some shady business. They were involved in illegal activities, and Kevin knew that he had to do something to stop them.

Kevin decided to hack into the company's website and post a warning to their customers. He wanted to let them know that the company was not trustworthy, and that they should avoid doing business with them. Kevin knew that his actions would get him in trouble, but he didn't care. He was doing what he thought was right, and he was willing to face the consequences.

The company was furious when they found out what Kevin had done. They threatened to sue him, and they even tried to have him arrested. But Kevin was not afraid. He knew that he had done the right thing, and he was willing to fight for what he believed in.

In the end, Kevin was able to expose the company's illegal activities, and he helped to protect the company's customers. He also learned a valuable lesson about the importance of doing what is right, even when it is difficult.

but isn't this "story" deeply boring?

Yes, but it seemed like the prompt asked for a "boring" story. I guess you could make it more interesting by writing a more detailed prompt

The prompt didn't ask for a boring story; nothing suggests that at all.

FREE KEVIN

Reply on news.ycombinator.com