Ask HN: Is GenAI heading for a crypto-esque bubble pop?

I wouldn’t compare GenAI to cryptocurrency because there’s no equivalent pyramid scheme mentality of “invest early in this asset and sell it later 100x” - it’s much closer to a traditional tech hype bubble. The average user isn’t roped in by the promise of buying and reselling a token, but by using and possibly paying for a product that can theoretically be useful to them on its own.

Like any tech bubble, you have crowds of investors pouncing on the “next big thing”, which will inevitably be overhyped and oversold, leading to the bubble popping when expectations outpace actual returns; but the core technology is still valid. The dotcom bubble popping didn’t mean the Internet was a fad, the video game crash of 1983 didn’t permanently doom the industry, the railway mania in the 1840s didn’t mean no one ever took a train after it crashed… any new and exciting development attracts grifters, but it eventually stabilises once people learn to separate the hype from the real applications of the tech. I remember when VR went through this - it was absolutely overhyped, but today there’s still a solid community around it even if it didn’t change the world.

With open source models and fine-tunes being uploaded on Huggingface, there’s no fraud, they’re perfectly free, and you can test them out and see for yourself that some work better than others even if benchmarks only tell part of the story. There’s absolutely a glut of thin veneers over APIs and models offering questionable benefits, but there’s also real value in having a solid UI on top of a basic model; we’ll see the ecosystem stabilise eventually.

> “invest early in this asset and sell it later 100x”

Literally VCs.

And many founders. The only difference to crypto here is that it's not the average retail investor doing pumping and dumping, it's the VCs and founders. Most founders will try to cash out with an acquisition while the hype is strong before the thing crashes and people realize there isn't any value behind it. Just like, you know, an ICO, but regulated.

Just to clarify, I'm not talking about LLMs in general, which I do believe have value, but about the current wave of low quality AI startups in the bubble.

This is always the case though with a new area of tech. The speed and churn of these "pump and dump" actions is slow, and very high effort (a typical acquisition will have a multi-year cliff requirement for founders).

Yeah, I'm just pushing back on the claim that crypto assets are fundamentally different. It's the same scheme, just faster and unregulated.

My two cents is that we will see a bubble pop, but it will rebound massively much like SaaS following the dotcom crash.

My primary concern is the ratio of AI-infrastructure products being built relative to the average utility delta of current end-user products. There are some shining counterexamples to this like Github copilot, but for the most part the utility gain from most text-based generative-AI seems incremental.

Current language models are bad to okay at most tasks - and great at a select few. The capabilities in unstructured ETL and production of labeled datasets fall in the latter category.

I think the next generation of really powerful capabilities to be unlocked with generative-AI will come with 1) an increase in compute availability (training even larger models) 2) advances in multimodal models, and 3) better interfaces to interact with them. Spatial computing will be a really big deal here, and people seem to have forgotten that Apple may have unlocked the next version of interactive computing earlier this year.

I'm not convinced Apple will crack the spatial computing or VR market. I think they have a polished product, but it looks like something that will be gathering dust in a few months if I bought one, sitting right next to my Quest Pro. At least the Pro can run a few games. I strongly anticipate that, unless you have a MacBook, you won't see the full potential of Apple's offering.

There was an article shared here around launch time. The quest pro is better seen as a dev prototype than a mass market product.

In my opinion the significant innovation is the design and implementation of the precision gesture control interface.

I can very much see that in use a decade from now with some future devise iterations be they laptop/watch/glasses/iPad.

#horselesscarriage

Forget about frameworks. Langchain is to GenAI what Doge was to crypto.

Most models on HF are like CrapCoins of the crypto era, so forget about 99.5% of them. There are only a few that can be useful (mostly +70b params).

Don't forget that some of this hype can be driven by nVidia to sell more GPUs. Some of it is driven by crypto enthusiasts who now use their GPUs for GenAI.

Forget about benchmarks and SOTA performance claims. Data leaks and you'll never be 100% sure that the team didn't mess with training data to manipulate the benchmarks.

Is GenAI a hype? Most likely, yes. Can it still be useful? Yes. You can still use GPT-4 and Llama 2 70b models for various tasks. Just don't use Langchain or any frameworks. Make your own API calls. And the smaller models (3-7b)? They are useful for when you don't want to chat with the model, like sentiment analysis.

Can you expand on why do you think langchain is so useless ? I'm quite happy using it for development.

Can you elaborate a little on why LangChain is bad and should be avoided?

I had modest success in using LLMs to achieve productivity improvements and results for my clients that are otherwise not possible. I used more direct API calls to invoke LLMs so far but am now considering switching to LangChain to make it more maintainable and scalble to turn more of my ideas into products and service offerings for my clients.

At the outset, a standard library that warps around popular LLMs and makes calling them easier seems needed and inevitable.

Low quality code and a lot of technical debt.

Okay -- I havent come across a critique of LangChain yet. Just starting to learn and use it.

I suppose if LangChain doesnt find a good path to becoming more mature -- then some other equivalent open-source library will replace it in coming months.

Thanks for the heads-up.

Can you substantiate this claim?

```

def embed_documents(self, texts: List[str]) -> List[List[float]]:

    texts = list(map(lambda x: x.replace("\n", " "), texts))
    embeddings = self.client.encode(texts, \*self.encode_kwargs)
    return embeddings.tolist()

```

lol when has this ever been an issue other than philosopher programmers

LangChain will not make the project any more maintainable than you can on your own. Avoid it.

Partially true...if you are looking for general purpose models. Instead of general purpose, I rather like more specialized models.

I have some transcription correction needs, and I had a set of prompts that could get GPT4 to do it...but boy was it expensive, and I had to break the problem down into multiple steps to get accurate results, increasing costs further.

My 13b Llama2 fine-tune Inkbot ( https://huggingface.co/Tostino/Inkbot-13B-8k-0.2 ) is able to do those same tasks with a single shot prompt. I have some examples of the specific tasks I trained that model on in the readme. But it isn't intended to be a general purpose model. It has a specific skill-set, and it's best to stick to those types of tasks. I'm also able to do batch inference at over 1k tokens/sec with this model on my desktop. That is something simply not possible with a 70b model on consumer hardware.

The reason we're seeing so many models right now is that the field is WIDE open for big leaps forward.

Prior to LLaMA in February there were almost no openly licensed language models that were any good at all - BLOOM and GPT-J were pretty much the best of the bunch.

Then LLaMA happened and the flood gates opened - there was now a base model which people could fine-tune against to get really great results.

Llama 2 accelerated things even more because it's licensed for commercial use - so now it's worth spending real money building on top of it.

Llama 2 isn't even 2.5 months old at this point. I'm not at all surprised at the rate of innovation around it - there is SO MUCH scope for improvement, and so many new techniques to be explored.

I expect we'll see a slow down in the release of models at some point, once the low hanging fruit has all been picked off - but it's going to be a while away yet.

You could even argue it's not so much the models themselves (LLM or diffusion), but rather the fact that fine tuning is remarkably accessible. Research is hot on ways to quantize parameters, swap portions for low memory, and the like. It's amazing to watch and use them.

I have yet to find LLMs useful beyond basic factoids (I have plans to deploy Llama 2 as an Alexa replacement), but the latest Stable Diffusion model (XL) has continued to impress me. It is clearly a game changer in the utility it provides to my interests. Admittedly, I'm having fun making high quality 8k wallpapers of abstract concepts and fictions places, but a truly honed creator could clearly work wonders with it.

Bitcoin is a solution in search of a much bigger set of problems than it can currently solve now. The ML Renaissance we've hit right now has clear, immediate utility to many. The moves made by the stock art heavyweights is proof enough of how disruptive it is.

Contrast this with how banks and financial institutions have responded to bitcoin. To my knowledge, there's yet to be a Bank of America Coin, a Wells Fargo blockchain solution, or non-fungible tokens offered by RBC. They clearly felt at least somewhat threatened, but not in any meaningful way.

Continued fun can lead to innovation and breakthroughs

LLMs are extremely useful for ML classification.

1. If you have a large amount of unlabeled data, labeling it using LLMs(even a costly XXXB parameter model) can be significantly cheaper than using humans.

2. Text classification works far better with LLMs compared to usual techniques like random forests.

3. We are exploring using LLMs on structured data(tables) for tasks like clustering, where it is difficult to tune other unsupervised approaches. Similarly we are exploring using LLM embeddings for similarity on unsupervised datasets.

4. LLMs could explain their decisions to users(e.g. why was your comment removed?) although work needs to be done on verifying it's correctness.

I wouldn't be surprised if LLM takes the spot of Random forest as the default go to for supervised ML.

Yup, and you can fine-tune an LLM to be a specialized classifier with relatively little data compared to building a classifier from scratch. I think that's the biggest benefit to LLMs, the data required to show it a behavior is pretty minuscule in the scheme of things.

My guess is that it’s heading for a late-1990s-Internet-esque bubble pop. Some overly hyped companies will go belly up and some investors will lose their shirts, but the technology will grow only more powerful and the world will be transformed by it.

No. LLMs are underhyped.

There are a bunch of models because it's easy to grab llama2 etc and do a little bit of fine tuning. There is nothing wrong or bubbly about that.

Drawing a comparison to cryptocurrency is shortsighted. Crypto, unfortunately had a very strong rival: Country-run fiat.

AI doesn’t have any rivals, perhaps faster and lower-power chips but we haven’t seen all the implications yet.

Surely human workers are rivals to AI.

I can write poems too ! Maybe not as variate as an AI, maybe not as fast, but at least I'll have an intent behind the words I choose, making the exercise funnier for my readers.

Humans have desires, and that make them sometimes way more interesting to listen to than AIs, until at least the AI has desire functions worth talking about. They also understand what the other side might have experienced, and give answers that will be more relevant to a human. For instance I asked "Is crypto a scam, fundamentally ?" to chatGPT and he gave me a well meaning very nuanced "no", synthetizing all sources. I would have answered myself: "it sounds useless, it feels dangerous, it may not be a sincere scam in all instances, but you should stay away if you don't have surplus money": am I not more useful ?

> am I not more useful ?

Yes, you are more useful to validate your own opinion and disregard ChatGPT as worse because it didn't agree with you. That's mainly the experience of people that fights against the tech in place of learning how to use it.

There is definitely a lot of hype and cherry pick demos showcased on social media, but the latest language models actually provide real productivity value unlike crypto. So it's not really fair to draw that comparison.

Personally, I haven't found a tool like ChatGPT+ that instantly improved my productivity overnight. Brainstorming, writing code, documentation, marketing copy, transforming data etc all became easier with this tool. I'd rather spend a shorter amount of time providing clear instructions to a language model than writing a lot these things from scratch.

Anecdotally, I've tried all the major models from Google, Meta, Anthropic, Stability etc nothing comes close to GPT-4 in understanding instructions properly.

What bubble? If you mean the hype bubble of every company trying to build LLMs into their product, then yes. IMO LLMs will render a lot of existing software obsolete or at least so fundamentally change the way we interface with them that the incumbents are likely to perish to newer, more innovative companies. If you mean a stock market bubble... I don't see one of those at the moment.

If you mean the bubble of excitement for AI, then I highly doubt it. LLMs are paradigm shift in how we use computers and what we can accomplish with them, and we've barely scratched the surface of them.

I don't think GenAI is going to fail to get off the start line, because it has done just that.

It has adoption that might not be every person in the world, but the depth and breadth and variety of users is hard to ignore.

Crypto wallets and getting up and running is the failure for most crypto. Yes, there have been improvements to that.

LLM's that took a chat interface went as universal as they could - just a computer anyone can talk to.

One issue with looking at HuggingFace and expecting maturity in 10-12 months is I'm not sure what can be compared as an equal in that number of months in other technologies. I agree it's far from perfect, but also it seems to be improving.

LLM's are just one kind of GenAI as you have pointed out. The AI you were working with before all of this remains relevant.

There's more to this than wanting a prompt to do everything magically. Having to "fine tune" and get each piece working may be the work for a while, and still be usable and beneficial.

Does it need to evolve and get better? It is.

With each passing month are more and more models running on more accessible hardware? It is.

Are people quietly doing a lot with it and not sharing it yet? Sure feels like it.

I actually think its the opposite, the next generation of models will rock the world.

Like with crypto before it, there's a tiny core of good stuff in genai, and an industry of rent-seekers adjacent to that using it for cover.

The competition for AI is simply any solution that gets a higher quality of answer than a trained approximation. That's actually a lot of things, but often there's a chain of dependencies in the information lifecycle that lead to combinations of technologies, and so our total solutions frontier expands by having the tech as a "drop-in" option that would otherwise be some custom engineered thing. But the AGI hype, to name one "unlikely AI thing", is silly. Based on what we know, larger genai models are likely to be more normal, not more novel. And normal isn't right or good, it just tells us what we want to hear.

Crypto is actually covering a polar-opposite technological context: it lets you get very high quality answers within a very narrow space. Done right, it tells you what you DON'T want to know.

In my opinion, which doesn't mean much, we're in a very weird spot where it's like 50/50 on usefulness and uselessness. There are certainly approaches where GenAI can be incredibly helpful, I've seen it in action in the day job. However, there are also avenues where it's complete snake oil.

I think applied AI, AI saas, and so on built on top of the best models to optimize specific workflows has a bright future. There will be failures and false starts as with any new technology, but there are already a number of useful apps out there with wide adoption, which is more than crypto ever achieved.

With crypto, the product is speculation. There are some potentially interesting applications for crypto, but it's never gotten large-scale traction as anything other than a form of gambling. AI doesn't have speculation built in like crypto does. It has to actually solve problems and do useful things to make money, and this is already happening.

There may be a speculative bubble in the VC realm, but when isn't there? It happens literally every time a new startup trend comes around.

Gen AI is merely one component of AI. Crypto and web3 are what you see what you get.

There’s still a lot of evolution to be had with AI; for example gen AI to interactive AI that uses all the bits we’ve seen so far to compose experiences with more than chatbot interaction.

If anything is going to die in AI it’s all the bespoke projects, not AI as a whole.

Am working on a Linux distro that ditches userspace as we know for a local LLM (root login is still available). The goal is no user accounts, /home and the AI encodes user data as vectors to disk. Need to successfully unlock the vectors rather than a user account.

So far it’s Linux From Scratch, Wayland, hyprland which displays the GPU viewport.

There is still a lot of weird stuff to do with AI. What do I use blockchain for except as yet another fiat currency

Could you please stop creating accounts for every few comments you post? We ban accounts that do that. This is in the site guidelines: https://news.ycombinator.com/newsguidelines.html.

You needn't use your real name, of course, but for HN to be a community, users need some identity for other users to relate to. Otherwise we may as well have no usernames and no community, and that would be a different kind of forum. https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...

I would say you are partially right. The number of people jumping on the AI bandwagon, hoping to get rich quick is incredible. In that, it resembles crypto, especially the NFT hype of a surprisingly short time ago.

Many of these companies exist solely to soak up VC money. They will take as much of the money as they can and then fail massively because they have no actual plan or no expertise with which to execute one. It might not qualify as fraud - or it might. I don't know.

So there will be a serious "adjustment" one of these days. Until then, keep your head on straight and be VERY careful.

Yes and no.

Generative AI as a technology isn't going anywhere and is very much going to increase in scope, scale, and necessity in most workflows.

But the vast majority of AI companies that cropped up riding the swell are going to be out of business in the near future.

Which is pretty standard. 90% of new companies aren't still in business a decade later no matter the field.

It's just that there's been a lot more new companies in a single new space since maybe 1999.

So expect a significant 'pop' of a bubble without there being any underlying issues with the tech or its applications.

(Possibly even more than typical given the rate of change and likelihood any custom product right now is building into its own obsolescence.)

Yes.

The downrounds and devaulations for the ChatGPT wrappers are coming in [0] and realizing that most of these 'AI startups' are overhyped. Especially the 'Chat with PDF' web apps and basic AI copywriting apps using AI APIs.

To suggest otherwise is pretty naive, and the large AI providers (OpenAI, Google, etc) do compete against their own partners.

[0] https://www.theinformation.com/articles/jasper-an-early-gene...

there are hardly any publicly traded AI companies or highly valued ones, compared to the multitude of tech companies and crypto companies. Crypto is way worse because average people lose money, not just institutional investors. AI companies make actual revenues from business activity, compared to crypto, which is just speculation mostly.

No, crypto is not mostly speculation, it's mostly money laundering.

Generative AI will at some point have to plateau, if it hasn't already. About the bubble pop, the (real) world is no different today than from when ChatGPT launched. This is exactly the same effect Bitcoin has had since the launch of its whitepaper.

No, it'll pop more like the dotcom bubble. Like, yeah it's a bit overhyped at the moment, but there actually is a fair bit of there there.

No, we're just getting started with GenAI. Crypto only affects a number of people and industries, GenAI affects a much larger proportion.

Yeah. as google said, openai has no moat. hope we get alot of competition in this area since there is so much censorship.

Deja vu. Is the internet a fad was a very common question in the early 2000s too

Yes.

mean crypto utility is basically zero. mean AI utility is decent.

Yes.

The fact that you can't type this post without using so many acronyms that only insiders can read it suggests yes.