5 ms·

I cannot wait for this deluge of AI products to be over

37 points by ilrwbwrkhv 4y ago

Almost every single one of them is a shitty wrapper on OpenAI or something of that sort and there are so many of them. Everywhere I look, I find just them. Waiting for it to dry out and go away.

tjr 4y ago
It is impossible for a potato knish to have the same gravitational pull as Earth, regardless of its size. The gravitational pull of an object depends on its mass, and even the largest potato knish would have nowhere near the mass of Earth. In fact, the knish would need to be many times larger than the observable universe to have the same mass as Earth. -- ChatGPT (GPT 3.5)
- iab 4y ago
  It's a salient point
  - nathants 4y ago
    No, it is not possible for a potato knish to have the same gravitational pull as Earth. The gravitational pull of an object is determined by its mass and the distance between it and another object. Earth's mass is approximately 5.97 x 10^24 kg, which gives it a significant gravitational force. A potato knish, on the other hand, is a small pastry with a mass that is many orders of magnitude smaller than that of Earth. In order for a potato knish to have the same gravitational pull as Earth, it would need to have a mass comparable to that of Earth. This is physically impossible for a pastry of that size. Increasing the mass of the knish to such an extent would also cause it to collapse under its own gravity and form a very different type of celestial body, and it would no longer be a potato knish. — gpt4
- mirekrusin 4y ago
  If you were locked in a box whole your life, your observable universe would be small as well!
moose_man 4y ago
It's the crypto bros. They jumped their scamming into AI. Wait for Logan Paul to put his ChatGPT app. =/ Honestly crypto people are plague on society.
- version_five 4y ago
  Agreed. I think we're seeing so much of it, especially company launches, because everyone knows they need to urgently try and get in before the bubble pops. Look at all the Lauch HNs there have been in the past bit. If you've got a startup that does logistics or something, you can probably fundraise whenever, subject to macroeconomics. If you made am NFT wallet or a chat GPT-powered dog sharing app, you had / have a tiny window to try and snag some dumb VC money before it's chasing the next thing.
- anonzzzies 4y ago
  Not the same though; AI is helping me while crypto never did. The stable diffusion stuff is really great for prototyping things at a bizarre speed. Instead of drawing vague things (I am terrible at it), I can now send the pro designers quite accurate depictions of what we want, without the iterations and terrible ‘not that colour blue’ meetings. That was hard and a terrible waste of time before. Don’t mind paying for that at all. And (chat)gpt and the apis help us replace humans we never wanted to hire in the first place, but we had no choice because writing software to replace them was not in the budget (enterprise companies…); now it’s so cheap that it is in the budget, like coffee is. Some tools are worthwhile, most are not; how is that different than other software dev? I would just not put the letters gpt in it and just make it do something people value and will pay for. Gpt has to be in there because the ‘products’ are generally crap and jittery and slow, so to ask money and get away with it, gpt is a ‘pre-alpha quality’ sticker somehow.
phil611 4y ago
could be as big as the Dotcom burst or a fad like web3. hard to tell
pasdechance 4y ago
agreed If the number of articles in my RSS feed that I need to mark as read isn't an indicator of wasted CO2, it is an indicator of how boring my day will be.
jnord 4y ago
And that is most likely what will happen but not before your hard-earned pension money has funded numerous failed AI ventures. Not too different from how fintech and crypto has played out.
- anonzzzies 4y ago
  > fintech and crypto has played out Played out? Those are still going strong. BTC is almost 30k again after ‘it was almost gone’, Altman is banging on Worldcoin, most Twitter promoted stuff I get is crypto while I follow no one in crypto (actually quite the opposite). Same for fintech; a slew of people I personally know from meet-ups just got millions for vague copy/paste ideas. How did it play out? Feels like it has many years to go still.
matrix_overload 4y ago
Per Gartner hype cycle [0], we are approaching the peak of inflated expectations for LLMs. There will be some disillusionment for sure, but some niches will truly benefit from the new tech (fingers crossed, not only advertisement). [0] https://en.wikipedia.org/wiki/Gartner_hype_cycle https://en.wikipedia.org/wiki/Gartner_hype_cycle
oars 4y ago
The difference between this and crypto is that users are already gaining tangible value and benefit from these AI products.
- kybernetyk 4y ago
  And what value would that be? Other than using ChatGPT to write silly love letters? (Or generate spam).
  - version_five 4y ago
    It's a hallmark of these scam technologies that people point to how their "already being used". Nobody talks that way about stuff that's actually being used. You see people talking about the use itself. Go back a few years and there were lists everywhere of what blockchain was "already being used" for.
  - linseed_213 4y ago
    I'm rusty with specific languages and skillsets I rarely use (e.g. I don't use CSS every day, and forget Python JSON/XML/IO interfaces). If I need to do something that requires those skills, it used to take me an hour to get up to remember how to do everything, and another hour to do the work. With ChatGPT it took ~15min as it handled all the boilerplate/basic stuff I was previously grabbing from documentation. It's not hard or impossible to do without it, but it did save me significant time. If you're an efficient engineer with photographic memory, it will not significantly improve your workflow. If you're an average generalist programmer who uses Stackoverflow more than once a day, GPT models will save you time.
brundolf 4y ago
We'll probably see more diversity soon (not just "wrappers of OpenAI"), but for AI and LLMs in general I think we're just getting started; this is something like 2008 for smartphone apps. I think it'll be a decade at least before this new market gets saturated and the wave dies down to make way for the next thing (and that's assuming the fundamental tech plateaus and doesn't lead to AGI) Best to buckle up and get familiar
andrewstuart 4y ago
You'll be awaiting a long time. Crypto/blockchai/NFT never once proved real value. AI has instantly provided deep and real value - certainly to me. I'm looking forward to wave after wave of ever better AI products from everyone from Apple to Amazon to all the IDE developers, the toy makers and just about every other sector of the economy. AI is already reshaping the way some things work and it's only months since the real revolution started. In technology you can reasonably expect that pretty much everything will fundamentally change. The most exciting thing is that it is like the first days of the web - back then all we could think of to use it for was transfer printed stuff onto. It took years until the technology was understood and applications developed that leveraged its inherent strengths. Stuff is coming with AI that will blow your mind ten times more than what we have seen so far, which is already mind blowing. AI is going to result in advances in other fields like medicine and energy - things that no-one thought possible in their wildest science fiction dreams. New fortunes will be made, existing fortunes will be lost or will diminish greatly. You may have to find something else to read if you want to get away from AI.
- thefz 4y ago
  I am still convinced that if you find enormous value in your work with ChatGPT, you are the one being substituted soon.
- version_five 4y ago
  I can't tell if this is a charicature or a real post - either way it definitely supports they hype bubble hypothesis. A completely empty fawning missive saying "this will change everything" with no reference or connection to anything. We could easily substitute deep learning, web3, etc without having to change it.
  - anonzzzies 4y ago
    Below in this thread I jotted down some actually value I get from this AI hype that actively makes me more money, more productive and yet have more time for other things. Didn’t have that 6 months ago.
  - 4y ago
moonchrome 4y ago
My biggest problem with this space is that if you have anything remotely valuable to be a startup - guess who has your model and is willing to outcompete you in the software market ? Unless it's something big tech wouldn't touch like those dating bots I keep hearing about.
JimmyRuska 4y ago
I watch startup pitches every week. There's a flood of money pouring in at the same time waiting to invest in a GPT-4 based tech that disrupts <X> segment of <Y> industry. Lots of people saying they've been trying to do something for years using NLP techniques, or specialized software, but chatGPT does it fine out of the box (eg loans, scheduling, customer support, customer recommendations). In general this is not going away. Language models are wrong and hallucinate every now and then, but they often get all the basic questions right, and it's clear they'll be more than suitable replacements to humans for many tasks. In general if people engage with these models enough maybe they'll collect enough tuning data to make giant expert systems to reduce error over time. The tipping point in LLMs may not be the tech itself, but being just good enough for mass engagement, where all your users are like mechanical turk training data Also people will keep finding ways to prune the LLMs to be tinier, once some Fabrice Bellard level genius makes them run super fast on cpu with lower memory they will be unavoidable.
- blazespin 4y ago
  Yeah, everyone is at a different stage of grief right now with respect to coming to terms with fact that vaunted human consciousness is not at the center of the universe like they thought it was.
charlierguo 4y ago
The deluge is quite something, isn't it. Over 40% of the latest YC batch are related to AI/ML: https://www.ignorance.ai/p/analyzing-over-100-ai-startups-from-yc https://www.ignorance.ai/p/analyzing-over-100-ai-startups-fr...
- version_five 4y ago
  You should submit that article. I'd be interested to see a discussion of YC's strategy to cram as many AI companies in as possible. People bring it up on the launch discussions but I don't think that's a good place.
  - charlierguo 4y ago
    Sure, just submitted! We'll see if it gets any traction.

37 points by ilrwbwrkhv 4y ago

Almost every single one of them is a shitty wrapper on OpenAI or something of that sort and there are so many of them. Everywhere I look, I find just them. Waiting for it to dry out and go away.

tjr 4y ago

It is impossible for a potato knish to have the same gravitational pull as Earth, regardless of its size. The gravitational pull of an object depends on its mass, and even the largest potato knish would have nowhere near the mass of Earth. In fact, the knish would need to be many times larger than the observable universe to have the same mass as Earth. -- ChatGPT (GPT 3.5)

iab 4y ago
It's a salient point
- nathants 4y ago
  No, it is not possible for a potato knish to have the same gravitational pull as Earth. The gravitational pull of an object is determined by its mass and the distance between it and another object. Earth's mass is approximately 5.97 x 10^24 kg, which gives it a significant gravitational force. A potato knish, on the other hand, is a small pastry with a mass that is many orders of magnitude smaller than that of Earth. In order for a potato knish to have the same gravitational pull as Earth, it would need to have a mass comparable to that of Earth. This is physically impossible for a pastry of that size. Increasing the mass of the knish to such an extent would also cause it to collapse under its own gravity and form a very different type of celestial body, and it would no longer be a potato knish. — gpt4
mirekrusin 4y ago
If you were locked in a box whole your life, your observable universe would be small as well!

moose_man 4y ago

It's the crypto bros. They jumped their scamming into AI. Wait for Logan Paul to put his ChatGPT app. =/ Honestly crypto people are plague on society.

version_five 4y ago
Agreed. I think we're seeing so much of it, especially company launches, because everyone knows they need to urgently try and get in before the bubble pops. Look at all the Lauch HNs there have been in the past bit. If you've got a startup that does logistics or something, you can probably fundraise whenever, subject to macroeconomics. If you made am NFT wallet or a chat GPT-powered dog sharing app, you had / have a tiny window to try and snag some dumb VC money before it's chasing the next thing.
anonzzzies 4y ago
Not the same though; AI is helping me while crypto never did. The stable diffusion stuff is really great for prototyping things at a bizarre speed. Instead of drawing vague things (I am terrible at it), I can now send the pro designers quite accurate depictions of what we want, without the iterations and terrible ‘not that colour blue’ meetings. That was hard and a terrible waste of time before. Don’t mind paying for that at all. And (chat)gpt and the apis help us replace humans we never wanted to hire in the first place, but we had no choice because writing software to replace them was not in the budget (enterprise companies…); now it’s so cheap that it is in the budget, like coffee is. Some tools are worthwhile, most are not; how is that different than other software dev? I would just not put the letters gpt in it and just make it do something people value and will pay for. Gpt has to be in there because the ‘products’ are generally crap and jittery and slow, so to ask money and get away with it, gpt is a ‘pre-alpha quality’ sticker somehow.

phil611 4y ago

could be as big as the Dotcom burst or a fad like web3. hard to tell

pasdechance 4y ago

agreed If the number of articles in my RSS feed that I need to mark as read isn't an indicator of wasted CO2, it is an indicator of how boring my day will be.

jnord 4y ago

And that is most likely what will happen but not before your hard-earned pension money has funded numerous failed AI ventures. Not too different from how fintech and crypto has played out.

anonzzzies 4y ago
> fintech and crypto has played out Played out? Those are still going strong. BTC is almost 30k again after ‘it was almost gone’, Altman is banging on Worldcoin, most Twitter promoted stuff I get is crypto while I follow no one in crypto (actually quite the opposite). Same for fintech; a slew of people I personally know from meet-ups just got millions for vague copy/paste ideas. How did it play out? Feels like it has many years to go still.

matrix_overload 4y ago

Per Gartner hype cycle [0], we are approaching the peak of inflated expectations for LLMs. There will be some disillusionment for sure, but some niches will truly benefit from the new tech (fingers crossed, not only advertisement). [0] https://en.wikipedia.org/wiki/Gartner_hype_cycle https://en.wikipedia.org/wiki/Gartner_hype_cycle

oars 4y ago

The difference between this and crypto is that users are already gaining tangible value and benefit from these AI products.

kybernetyk 4y ago
And what value would that be? Other than using ChatGPT to write silly love letters? (Or generate spam).
- version_five 4y ago
  It's a hallmark of these scam technologies that people point to how their "already being used". Nobody talks that way about stuff that's actually being used. You see people talking about the use itself. Go back a few years and there were lists everywhere of what blockchain was "already being used" for.
- linseed_213 4y ago
  I'm rusty with specific languages and skillsets I rarely use (e.g. I don't use CSS every day, and forget Python JSON/XML/IO interfaces). If I need to do something that requires those skills, it used to take me an hour to get up to remember how to do everything, and another hour to do the work. With ChatGPT it took ~15min as it handled all the boilerplate/basic stuff I was previously grabbing from documentation. It's not hard or impossible to do without it, but it did save me significant time. If you're an efficient engineer with photographic memory, it will not significantly improve your workflow. If you're an average generalist programmer who uses Stackoverflow more than once a day, GPT models will save you time.

brundolf 4y ago

We'll probably see more diversity soon (not just "wrappers of OpenAI"), but for AI and LLMs in general I think we're just getting started; this is something like 2008 for smartphone apps. I think it'll be a decade at least before this new market gets saturated and the wave dies down to make way for the next thing (and that's assuming the fundamental tech plateaus and doesn't lead to AGI) Best to buckle up and get familiar

andrewstuart 4y ago

You'll be awaiting a long time. Crypto/blockchai/NFT never once proved real value. AI has instantly provided deep and real value - certainly to me. I'm looking forward to wave after wave of ever better AI products from everyone from Apple to Amazon to all the IDE developers, the toy makers and just about every other sector of the economy. AI is already reshaping the way some things work and it's only months since the real revolution started. In technology you can reasonably expect that pretty much everything will fundamentally change. The most exciting thing is that it is like the first days of the web - back then all we could think of to use it for was transfer printed stuff onto. It took years until the technology was understood and applications developed that leveraged its inherent strengths. Stuff is coming with AI that will blow your mind ten times more than what we have seen so far, which is already mind blowing. AI is going to result in advances in other fields like medicine and energy - things that no-one thought possible in their wildest science fiction dreams. New fortunes will be made, existing fortunes will be lost or will diminish greatly. You may have to find something else to read if you want to get away from AI.

thefz 4y ago
I am still convinced that if you find enormous value in your work with ChatGPT, you are the one being substituted soon.
version_five 4y ago
I can't tell if this is a charicature or a real post - either way it definitely supports they hype bubble hypothesis. A completely empty fawning missive saying "this will change everything" with no reference or connection to anything. We could easily substitute deep learning, web3, etc without having to change it.
- anonzzzies 4y ago
  Below in this thread I jotted down some actually value I get from this AI hype that actively makes me more money, more productive and yet have more time for other things. Didn’t have that 6 months ago.
- 4y ago

moonchrome 4y ago

My biggest problem with this space is that if you have anything remotely valuable to be a startup - guess who has your model and is willing to outcompete you in the software market ? Unless it's something big tech wouldn't touch like those dating bots I keep hearing about.

JimmyRuska 4y ago

I watch startup pitches every week. There's a flood of money pouring in at the same time waiting to invest in a GPT-4 based tech that disrupts <X> segment of <Y> industry. Lots of people saying they've been trying to do something for years using NLP techniques, or specialized software, but chatGPT does it fine out of the box (eg loans, scheduling, customer support, customer recommendations). In general this is not going away. Language models are wrong and hallucinate every now and then, but they often get all the basic questions right, and it's clear they'll be more than suitable replacements to humans for many tasks. In general if people engage with these models enough maybe they'll collect enough tuning data to make giant expert systems to reduce error over time. The tipping point in LLMs may not be the tech itself, but being just good enough for mass engagement, where all your users are like mechanical turk training data Also people will keep finding ways to prune the LLMs to be tinier, once some Fabrice Bellard level genius makes them run super fast on cpu with lower memory they will be unavoidable.

blazespin 4y ago
Yeah, everyone is at a different stage of grief right now with respect to coming to terms with fact that vaunted human consciousness is not at the center of the universe like they thought it was.

charlierguo 4y ago

The deluge is quite something, isn't it. Over 40% of the latest YC batch are related to AI/ML: https://www.ignorance.ai/p/analyzing-over-100-ai-startups-from-yc https://www.ignorance.ai/p/analyzing-over-100-ai-startups-fr...

version_five 4y ago
You should submit that article. I'd be interested to see a discussion of YC's strategy to cram as many AI companies in as possible. People bring it up on the launch discussions but I don't think that's a good place.
- charlierguo 4y ago
  Sure, just submitted! We'll see if it gets any traction.