Besides the science behind it, currently it feels like the same hype as crypto a couple years before.
A week after ChatGPT was out and plenty of people were already using it for writing code, emails, and plenty of other tasks. It would be weird to argue that AI is not going to have a massive impact at many levels.
Now there are plenty of challenges to overcome of course, but I have no doubts that something that useful on day one is going to have a big impact once we really understand how to integrate it to various products.
There WILL be job losses but it's not going to be the kind of people who hang out on HN. I can't think of any reason why you wouldn't have an AI taking orders at the drive through, handling customer service calls / tech support, etc. Any job that consists mostly of having the same simple, repetitive conversations is going to eventually be cheaper to have a computer do.
Same thing is happening with “AI”. I think it’s not a fad but at the same time it is.
In the company I work for, in the media sector, there’s clear and direct use of LLMs (it’s already being done, and yes it will mean quite a few people will lose their jobs, despite many HNers saying that it won’t happen), but with all the hype they want to get some sort of “AI” everywhere and the most ridiculous ideas are being POC’ed.
The stable coins are an issue but stable coins aren't a necessity for btc transactions, they're a utility for traders. Btc's biggest danger is its price volatility. As it increases in value and market cap it will become more valuable to large players who will in turn be motivated to protect its integrity.
Same for developed countries, though on a small scale.
In practice, dedollarisarion has been about shifting to other major national currencies, like the yuan. Bitcoin doesn’t really feature.
If btc replaces 1% of fx it's doing $50B/day of transactions. Market cap of btc is only about $600M, so 100x btc is not crazy.
Is a p2p transaction system with an auditable record attractive to 1% of fx transaction parties? Seems reasonable, especially when corrupt states are a counterparty.
how does the integrity of your counterparty to a transaction affect the currency or medium of exchange you agree to?
> a) accept their local currency? where? what then? what will it be worth by the time we exchange it? > b) take btc. transaction registers publicly on the blockchain. smart contracts can execute if desired, e.g. when payment x is received to address y initiate a sequence that starts delivery. currency risk is now in btc and that can be instantly mitigated by converting to your chosen currency, as the btc market is liquid. > c) require counterparty to pay in your currency, which means they pay exchange fees, increasing their costs.
Isn't it $600000M ?
>e best of tulips cost upwards of $1 million in today’s money (but with many bulbs trading in the $50,000–$150,000 range
https://www.investopedia.com/terms/d/dutch_tulip_bulb_market...
But soon after they realized it wasn't as useful for these tasks as initially thought. Interestingly, a lot people started to believe that the tool had been limited on purpose, where in fact they were just becoming a bit more objective.
That being said, it's better than crypto and there are applications (although maybe not life changing). The bar is low though.
Currently, nobody is talking about ChatGPT more than the crypto/NFT hustlers who now need a new angle.
Most of the actual techies in my circles, as in: practicing software engineer for 20 years kind of people, looked at it, maybe tried it out of curiosity, and went back to work.
The people who are really into it are the same people who were really into SEO, and then leadgen marketing, briefly online poker, and then blockchain/crypto, and then NFTs: the eternal hustlers who just look for the next hype train and ride it until the next train enters the station...
The interesting difference with AI / LLMs is that for whatever reason, the big companies have fallen under the spell, and they're all trying to cram generative AI into all their products now. I work at one of these companies and it's bizarre how they've turned the huge ship on a dime and are now trying to AI All The Things.
My moment was when I realized if you ask ChatGPT a question about itself, like how ChatGPT works, you are not receiving an authoritative or 1st person kind of answer, the way everyone assumes. You are getting a rehash of press releases from a text autocomplete engine. Everyone when they interact with it intuitively feels they are receiving an authentic, slightly flawed, interaction with intelligence, but it's just PT Barnum with a text completion feature. Bravo.
That’s how most human beings work as well. Ask someone about what India or Thailand is like, and even if they’ve never been there, they’ll be happy to give you a rehash of stories, pictures, and videos they saw about India or Thailand. They might make up totally wrong facts as well, just like ChatGPT.
Citation needed.
I put it in the same bucket as deep CNNs - very good for specific tasks, but ultimately their lasting impact will be something trivial and faily non-world-changing like being able to search your photo collection in a slightly more clever way.
I keep seeing this take and it doesn't make any sense to me. Some tech has obvious utility and some doesn't.
For example, I knew internet (web, email) were valuable when I discovered them because accessing information and communicating with people were already things I did - the internet unequivocally made them faster and easier and often cheaper.
chatGPT / bard gave me a similar vibe - I use them to brain storm/shape ideas, and as mentioned elsewhere, they do a great job of tasks like drafting a job ad. These are things people already do and this tech just makes it better. So people will use it.
In contrast, I "get" why people were excited by crypto but I don't personally know anyone whose payment/banking experience is improved by it. As an American for example there was nothing tangible Bitcoin made easier vs my bank account and visa. So it was always less "obvious" that it was going to be a valuable thing beyond hype.
So far the only way this even comes close to happening is by wrapping centralised systems around it (e.g., exchanges), but that comes with a whole set of different drawbacks. This does not feel like a scenario that is possible, let alone likely.
That's exactly because you live in a first world country with a reliable banking system.
My take is that, for most non-tech people, this is their first experience directing a computer to perform a precise task - ie programming. They're accustomed to using applications, not having a hand in making them. Maybe they've used Excel before and felt a bit of this power. But ChatGPT allows them to dream up a novel idea and get the computer to execute it - something that feels like magic to non-tech folks but is rather pedestrian for most of us in tech.
It's actively solving problems for people and saving time / creating efficiency. That has value and economic utility.
I honestly don't get what a lot of the cynics in this thread and their "highly technical/IT friends" are missing.
But does it enhance the end product? Does it improve upon the work of already competent developers? Does it actually solve the real problems that software engineers face? Highly debatable. It certainly makes cranking out code more "efficient" - but anyone who's every created software knows that cranking out lines of code is a terrible metric for success. Poor quality code has less than zero value, it's a liability. As the prevalence of bot-generated code goes up, it will place an increasingly high burden on actual professionals to clean up the mess.
It is quite a stretch to say that lay people can effortlessly write application only with AI whatever it is copilot or chatgpt.
You still have to know a lot of things to build simple crud to store data about stamp collection.
LLMs are not changing that.
But well over a decade ago, we had high-level frameworks for translating "models" into all the gritty details of a crud app. Model description goes in. API endpoints, database migrations, admin UI, all auto-generated. I even developed something like it based on Django 0.96 (now abandoned).
It's not a stretch of the imagination to generate your model.py with model.txt containing the problem description in plain text.
My experience with every one of them is that getting the model to the point where it will generate something useful is more work than just writing the useful part myself.
Your second point re: how model-driven code generation can be a dead-end... I totally agree. I didn't say it was a good idea! But it was motivated by a real need: generating lots of web apps in a particular domain.
In other words, the tech is real and amazing, but it doesn't seem as immediately useful in the short term as some people expect.
Same, and they are using it. Tech tends to look at the exception cases or try to use it for exact answer types of things. But non-tech are happily using it as they would Google to come up with ideas for parties and events, write the mundane emails many people have to send, etc... I've been using it to bounce ideas off of and build things like marketing plans. Basically dynamic templates. The challenge right now is prompt engineering.
Could this perhaps be a you issue? How do you feel when you think about people changing from microservice architecture to blockchain to crypto and now to language models?
This feels like a common pattern; a few years back many of my non-tech friends believed self-driving cars were coming imminently, whereas ~no-one working in tech believed that.
Company execs sure did. "Autonomous vehicles" is the mirage that Uber, Waymo and others dangled in front of investors and press for years. Can't blame the public for not figuring out that it was a fig leaf to paper over their eye-watering losses.
Upton Sinclair probably applies too:
> it is difficult to get a man to understand something, when his salary depends on his not understanding it
If you're Uber leadership, well, you're strongly incentivised to believe that there is _some_ way out, even if that way out might seem pretty implausible to a dispassionate observer.
Of course, one tends to be optimistic of these things when they work on them.
5 years ago I was hearing 10-20 years. This was from people who were connected and knowledgeable in the industry. Now the same people are singing a different tune.
Maybe in 20 years from now, but it's looking pretty impossible in the next 5-10 years.
The issue is that misestimation of a few key factors and overestimation of our current capabilities. If we already have the tech for self driving cars, maybe it will only take 10 years.
Assuming this something computers do quite well (wrong assumption, but it seemed reasonable at the time), we already have vehicles that can navigate themselves (we do), and we're only somewhat recently getting to the point that cars are sophisticated computers (mostly true, although computer control of car isn't something new.
There have been lots of ECUs for sometime, but the capabilities have really exploded. Maybe we just haven't gotten around to do self driving and the tools are in our hands right now!
When you understand how the auto industry works, 10 years is a relatively short timespan, that will result in ~2-5ish major design iterations. If you don't have the feature presently in the pipeline, the clock is really ticking, assuming it's new technology.
To deliver on the 10 year deadline, you will only have a couple years to be almost completely ready. 10 years is not "who knows", it's a bold prediction that implies the final product is imminent. People in tech didn't know what the fuck was going on, thinking we'd have taught the cars to drive and got the tech figured out in 5-7 years. It was loony toons.
Many very smart people I know, at least one of them inside the industry, seemed extremely confident of this. Things look different now.
It's just cargo culting and wishful thinking. People staring into the magic mirror, hoping it will clone their desires.
The phrase cargo culting or some meaningful equivalent should make a come back. Even after the 'great youtubening' of technical knowledge, it is easy to find people stuck in habituated imitation of tech skills and tech talk.
I love that people are interested, but cargo culting is no good water to drink from.
I absolutely do not believe this to be the case. For starters the average person probably isn’t even aware, but the vast majority of folks I’ve seen use it have found it super interesting for like a week then dropped it because it wasn’t actually more helpful/better than the previously available tools.
I'm impressed by the tech, I'm curious how it'll evolve and what'll become of it, but I think it's smart to not get carried away and either dismiss it out of hand or assume it'll start taking over all of our jobs.
That being said, I can understand why people dislike it, same way that my grandparents only use cash, they don't trust bank cards. Imagine if we still only used cash?
I'm sure there's some niche edge cases I'm missing, but where the 'post cash'/bank cards world has real advantages, how would cryptocurrency improve any normal persons day to day? I see no real advantage, it all seems more "you could also do this with crypto" - except maybe not as well/fast/cheaply/securely?
FWIW I don't hate crypto at all, I just can't see it becoming 'the thing'. AI I can definitely see having a place (already). I find ChatGPT insanely useful for some things.
> How would cryptocurrency improve any normal persons day to day?
I see your point, but I can think of at least a few situations:
1. Global transactions with lower fees
2. Remittance, if you're working abroad etc, no more WesternUnion fees
3. Alternative to cash for previously cash transactions (less chance of being mugged etc)
There are definitely arguments for it, but anything you can do with it you can do without it, so it doesn't invent new use cases really, but I find the arguments against it are never based on pragmatism and logic, but on emotion and bias.
Crypto's only true significant/impactful use case in finance is money laundering and facilitation of other types of crime.
Digital event tickets could be sold directly from the venue to the attendee without any of the hassle of TM. Tickets that are non-fungible (ie an assigned seat at an event) can be represented with NFTs. Fungible tickets (ie no assigned seats, maybe access to any event within a period) can be simply tokenized.
That isn't a use case that exists now, and because of the toxic association that the term "NFT" has, it probably won't exist. It's a shame though, because the technology exists, and works better than existing options.
I have bought assigned seat tickets online direct from venues without TM nor NFTs, worked fine.
I have also bought tickets to exhibitions (no seats nor assigned time) direct from venues online without TM nor NFTs, worked fine.
Seems that the only problem is TM and you won't make them go way with NFTs.
There is no fundamental difference if you swap the TM brand for another. By using an NFT-based solution, the middleman can be cut out entirely, while maintaining feature parity, and without introducing extra burden on venues. An OTS Free Software application can expose the same functionality to venues without needing to maintain any infrastructure, and without dealing with TM.
The Bitcoin paper was published in 2008. It’s been almost 15 years now - plus all of the advantages that you have with modern development practices that were unavailable in the early 2000s - and yet still nobody uses blockchain at scale for this use case.
If this worked better than existing options at a lower cost, then businesses would be rushing to adopt it.
The reasons for NFTs not being used for this use-case are way more nuanced then "it should have found success in the exact same number of years as selling tickets on the web".
And venues could do that (they’re not limited to selling tickets via Ticketmaster), but then they have to maintain technical and operational expertise which is a distraction.
Sounds like you've spotted a business opportunity!
Electronic money, instead of paper or coins, is there to stay. However, bitcoin is a very poor implementation of this. The mining alone has contributed significantly to the earth's yearly electricity consumption.
There are tons of legitimate use cases. However, most of them aren’t immediately apparent to affluent people in the United States as it solves problems most of them don’t have.
Are you implying that you cannot buy a car on a weekend in the US?
If you Google, you’ll see that not only can you buy a car on the weekend, but there are articles from insurance companies on adding car insurance, articles about whether it is better to buy on a week day or weekend, and even articles on buying on a 3-day holiday weekend!
the beautiful thing about accounting and trade is that cash flow doesn't have to coincide with the transaction.
You may not have meant it but it sounds like an argument made in bad faith. Cash has real world application and can be used anywhere. Not trusting bank card is a choice but the alternative can be used anywhere which is cash.
Crypto at best is used for get rich quick schemes, scams, dark web payments, ransom demands etc. You said buying pizza with bitcoin. Where ? Is it an anecdote ?
To unpack this:
1. Crypto at best is used for get rich quick schemes, scams, dark web payments, ransom demands etc.
I think this is the bad faith in this discussion. Cash is also best for scams, crime, ransom demands etc. Knives are used to stab people, so should we limit cutlery to spnorkfs?
2. You said buying pizza with bitcoin. Where ? Is it an anecdote ?
https://www.coindesk.com/consensus-magazine/2023/05/22/celeb...
While some thing might work, it doesn’t imply that it is practical.
My family who are mostly non-technical have heard of ChatGPT but couldn't tell you what it is or does.
In tech I see a wide spectrum of skepticism, with many people using it well and getting a lot of value out of it, and many remaining skeptical or having not found a good way to integrate it into their workflow yet.
This is obviously not true. There was a lot of skepticism within the tech world but also a LOT of hype and hubris, even coming from some of the most important (even if not most credible) voices in the industry like Andreessen.
> of the people outside tech, it was only the people who were into crypto (etc) who are into AI.
I hoped this would be clear, especially given that the next paragraph talks about my thoughts on people in tech. Apologies if it wasn't clear enough.
To their credit, they said, "wtf would we want to do that?!"
Which part of the industry are you working in?
I work for Google and hear a lot of talk about LLMs from my serious colleagues.
LLMs don't know what's right automagically, a model fine tuned by a hedge fund is definitely better than just an LLM's hallucinating something