7 ms·

Over 40% of the latest YC batch are AI/ML startups

63 points by charlierguo 4y ago

borissk 4y ago
The companies that stuck with paper and didn't adopt computers, the web and mobile went out of business. The same will happen in the near future with the companies that are unwilling or unable to adopt AI/ML in their business.
- palata 4y ago
  Survivorship bias. You select technologies that won, and assume your new technology will be the same. What about all those technologies that just died? Companies that did not embrace Minitel, blockchain or the metaverse seem just fine right now.
- bilqis 4y ago
  Typical grifter take, using analogies, which are always a bad framework to analyse and predict events and trends based on the history. A few years ago others grifters were saying the same about blockchain and NFTs, a year ago same thing about metaverse/decentraland.
  - borissk 4y ago
    Do tell - how do you analyse and predict events and trends NOT based on the history.
    - palata 4y ago
      A first good step IMO is to accept that we suck at predicting.
      - borissk 4y ago
        We are great at predicting. The number one task of the huge human neocortex is predicting the future. If you look at the past HN comments there were users who predicted the raise of OpenAI, the war in Ukraine and so on.
        palata 4y ago
        Did you forget a "/s", or do you actually believe that?
        SideburnsOfDoom 4y ago
        No, but they are playing word games. yes, the human brain is a great engine designed for predicting future events and working with that e.g. how to catch a ball that is coming towards you and will arrive shortly in the future, how to throw an object and hit a target, how to pick up a heavy object and move it to where you need it, etc. But this is not at all the same kind of "predicting future events" as accurately saying where e.g. the CryptoCurrency or VR industry will be in a decade or so.
        mym1990 4y ago
        Predicting isn’t some black and white skill. The neocortex doing calculations in the background to predict how heavy a cup of coffee might be or how hot the stove is…is very different from consciously predicting geopolitical events which are based on countless other variables. In a world of 8 billion people of course you will see some get things right, and then conveniently point it out…but most people’s predictions are wrong, and no one pays attention to those results.
    - SideburnsOfDoom 4y ago
      I'm puzzled by this take: surely the recent history - the cryptocurrency bubble - the hot air and lack of substance, the extractive scams, is useful lens for current events? But it's not everything, we can't be entirely "based on it" deterministically - as we do indeed "suck at predicting".
    - mym1990 4y ago
      That’s easy, if you have a framework of knowledge around things like physics or chemistry or biology(add any domain here), one could predict outcomes of hypothetical use cases with considerable accuracy. Unless you lump all domain knowledge into “history” because it has to have happened in the past…
      - Idiot_in_Vain 4y ago
        Do you think chemistry or biology knowedge was given to humans by the gods or the ancient aliens and not based on a history of practical knowedge, experiments and research?
        mym1990 4y ago
        Thats why I added the last sentence. But for the purpose of debate, it is reasonable to distinguish domains like physics from history. It is one thing to study how something was discovered, and another to understand the parameters around an event so that you can predict what might happen.
    - bilqis 4y ago
      My point was, that saying “historical event X is somehow similar to ongoing event Y, thus we can reason and predict outcomes of Y based on X” is a false methodology, appealing, but false.
  - nikanj 4y ago
    Remember when people paid good money for Second Life real estate?
  - SideburnsOfDoom 4y ago
    Exactly. What was the corresponding percentage of Blockchain and/or Cryptocurrency start-ups say, 2 years ago?
- Beaver117 4y ago
  Except these companies are just thin wrappers over OpenAI APIs
  - borissk 4y ago
    Today they may be thin wrappers over OpenAI APIs, but if they start bringing in big money they can hire data scientists and build their own models.
    - rvz 4y ago
      That is a big "if", which doesn't cut it with startups looking for VCs, given 95% of these ML startups depend on OpenAI's APIs. What I am looking for is 'even if'. So even if they make money AND hire 'data scientists', there is a higher chance that they will run out of money quicker since they are competing against Google, Microsoft and OpenAI which they all can afford to run everything at a loss for free and raise their prices which eats these startup's revenue. Either way, they are locked into these APIs and they aren't focused on profitability to even begin building their own models at this stage. If they are not profitable by the time the price increases come in, they are going to be struggling for survival.
- dddw 4y ago
  Frakkin' toasters everywhere!
- kromem 4y ago
  Yes, but there's a massive difference between a startup focused on a broader business issue incorporating AI into its plan on tackling that issue and a company focused on AI as the primary product. The former are going to mostly be successful, particularly in the place of companies that don't adapt, but the latter are mostly going to be obsolete by the time they'll be several years into it. AI as a product is a TERRIBLE play. But AI as part of a solution is going to be necessary for nearly every business vertical.
- onion2k 4y ago
  Some businesses absolutely have adopted IT with vigour and massively improved the way they work, but many haven't. They've just 'ported' their old paper processes to computers without learning or improving anything. Sending emails is faster than using paper memos but if you're just sending memos by email nothing is actually better. Using an Excel spreadsheet with no macros, formatting, or functions doesn't really give you much that double entry bookkeeping in a real book wouldn't already have given you. Plenty of people use Word with everything 'annoying' switched off, and they could do the same job on a typewriter. Similarly, lots of companies spent thousands of dollars on a website or a mobile app that no one looks at because they're a small town landscaping business that really doesn't need those things. The same goes for AI. There are definitely use cases where it will improve how we do things. For some businesses that will make them more profitable and more competitive. A lot of businesses won't adopt it for decades. Some will adopt it when it brings them no real benefits. AI is a tool. People will often use it poorly, just like every other tool.
- choeger 4y ago
  I have yet to encounter a single valuable (in terms of hard currency) use case for GPT besides automating publishing. Yes, you can generate text automatically now in relatively high quality. But that's a business that was struggling before GPT came along and I don't think we actually need that much automatically generated bullshit content. Yes, a language model will help authors to write their articles/books quicker, but that's not a significant fraction of their time spent, hopefully. I think language models will become a commodity, like a spell checker is today. They might open up the market for translations into more exotic languages (but I don't know how the quality is for such languages and by definition, the market ain't big) and I think every author will use one as part of their workflow (think of technical writers using a language model for the textual parts). Computer games might profit massively in terms of immersion. No, I don't think that new applications will become that important with language models. I think it's the other way around. There are so many applications that we simply don't need any longer. Language models and image recognition models together will take over the user interface industry. Why bother with web design and user experience if you can just throw the chatbot at your client? That's going to become the real value here. If I was a junior web developer, I would begin to seek out new programming languages and stacks.
wallaBBB 4y ago
This is not going to end well.
- warent 4y ago
  In what way?
  - onion2k 4y ago
    One possible way it could end badly would be if OpenAI ramped it's pricing up to the point where it takes away most of those company's chances of ever being profitable. I'm not saying it's going to happen, and I hope it doesn't, but when I did a little analysis for an idea I have it was absolutely the biggest threat I could see. OpenAI has the ultimate vendor lock-in - there literally isn't an alternative right now. (This is a giant opportunity if you're a billionaire investor. Providing an AI backend is probably the next cloud infrastructure. OpenAI is going to be the AWS of that industry, but there's room for an Azure or a GCP too.)
    - Mizoguchi 4y ago
      There's really no point in worrying about OpenAI shutting down their API or pricing you out. It may happen but very unlikely that it will be suddenly and soon. Yes, keep it in the back of your mind and design your systems in such way that when the time comes replacing the GPT doesn't require rewriting your entire codebase. Your immediate #1 priority in such competitive market should be to execute your idea and get paying customers, then you deal with breaking off OpenAI.
    - whiplash451 4y ago
      OpenAI’s moat is not _that_ big in the grand scheme of things. You can train GPT2 in a couple weeks on a dozen GPUs. Sure, it’s a lot worse than GPT4, but not insanely behind and the community is fast. Same for RLHF. It is for sure fine engineering, but not unreachable for a motivated and competent bunch.
  - wallaBBB 4y ago
derelicta 4y ago
oh yeah I've noticed this trend on indiehacker. Lotsa lotsa new products being advertised as AI-powered and whatnot. So it doesnt suprise me a lot.
fortissimohn 4y ago
Surely this technology will be used for the benefit of humanity. By 2030 we will be working 15 hour weeks and all people will have homes!
- tuyguntn 4y ago
  (for the sake of critical thinking) or rich get richer, poor get poorer and will not have access to AI tools, society will collapse, more jobs will be low paying, most on demand skills would be everything done by physical activity
  - admissionsguy 4y ago
    > society will collapse, The great promise of robotics and AI is the possibility of decoupling societal stability from the vagaries of the poor.
    - borissk 4y ago
      How? By having armies of semi-intelligent armed drones, that will kill the poor if they decide to riot?
      - kolinko 4y ago
        By bringing wealth so cheaply that it can be given away. We already have many things that are given away for free, whereas a 100 years ago they were considered luxury (books, pens, papers, access to knowledge in general, and in Europe - public transport and healthcare, plus to some extent food, heat and electricity).
        whiplash451 4y ago
        Then that’s not wealth anymore, because wealth is relative.
        Idiot_in_Vain 4y ago
        Would the wealthy and the powerfull be happy for everyone to become wealthy?
raverbashing 4y ago
How many of those are just a minor prompt + API call to OpenAI?
- yieldcrv 4y ago
  Look, you’re not going to get investor money to train an LLM and scrape a ton of data sources at this stage Is what it is
- mirekrusin 4y ago
  Probably more than 90%?
- onion2k 4y ago
  The challenge of building, running, and growing a startup is rarely building some novel algorithm or complex machine. You can reduce almost all SaaS businesses down to "It's just a UI and a database!" if you want to, and many developers do. That only shows how easy it is to over-simplify things. My first startup was literally just some Javascript and a Postgres database. It didn't even call any external services, let alone one that's truly on the bleeding edge of computer science. I still thought it was a viable business (wrongly, it turned out, but still... I had a dream!).
  - raverbashing 4y ago
    I find it funny some people being dismissive of what they built to defend the AI startups "Some JavaScript and a DB" is a web service. It does stuff. Sure if it's small or a basic crud then the moat is small But if I can replace the value proposition of a startup with "open chatgpt and type a short prompt" then there's not much value in it right. Maybe if there's value in the surrounding services.
    - onion2k 4y ago
      If the business is literally just sending data to GPT and outputting the response then sure, but I've not seen any businesses trying that yet. Pretty much everything I've seen to date is "A CRUD app with GPT to fill in blank data". That's automatically more valuable than just a CRUD app.
maz1b 4y ago
Surprisingly, there was only one (!) education startup in this batch. Doesn't bode well for us as we're applying in this Summer 23 cycle albeit with customers, revenue, PMF, etc.
- palata 4y ago
  [flagged]
  - UncleEntity 4y ago
    “AI education” if you really want to get the money. Somebody has to teach the AI how to do things.
- dang 4y ago
  Don't worry about that at all! It will help you stand out.
danielmarkbruce 4y ago
Not surprising. You can sit down and bust out something pretty cool in an afternoon using the infrastructure currently in place (openai, huggingface etc). Makes more sense than being the 57th SaaS web app in any given space.
- Beaver117 4y ago
  If the bar for entry is that low how will these companies survive?
  - palata 4y ago
    The probably mostly won't. But they don't care now. Now they want VC money. That's how startups work.
  - nikanj 4y ago
    Every investment fund is under strict orders to plow money into AI, because it's going to change the world
  - surgical_fire 4y ago
    They only need to survive for a long enough time until they find an exit. Either through acquisition or IPO. It doesn't need to be sustainable nor particularly innovative.
    - lma21 4y ago
      It's going to be tough to IPO when you're not sustainable or even profitable these days.
  - kolinko 4y ago
    That's where the stat that 99% startups fail comes from. One or few winners, a hundred losers - the whole startup ecosystem works like this :)
  - danielmarkbruce 4y ago
    The bar to making a social network weren't high from a technology perspective. Same for a lot of enterprise software.
ookblah 4y ago
feel kind of sorry (or excited?) for this batch lol. i have trouble believing that many companies were already working on AI/ML stuff. basically huge land/cash grab right now. surely at this level there has to be a lot of overlap... it really feels like just throwing everyone into the pit and having them duke it out and see who wins, loser be damned. i suppose that's the VC play in general, just a bit jarring to see it front and center.
alfiedotwtf 4y ago
"We're So Lo Mo... I mean So Mo Lo"
cornholio 4y ago
And most of those "companies" are shims over GPT. Zero moat, fake it till you make it mentality, gather investor money and maybe we'll roll our own in good time. Don't know how I feel about that, it quacks like a bubble but there is massive value that can be delivered.
- whiplash451 4y ago
  Sure. But as a VC you will look like a schmuck if you don’t invest in at least some AI startup.
- nikanj 4y ago
  It's the model familiar from previous bubble rounds: 1) Get funding 2) Put a lot of effort into appearances and presence 3) Get acquihired by Google / Microsoft / et al.
  - whiplash451 4y ago
    Except step 3 is going to be a lot harder these days. You’d better have a strong story to sell something to Satya.
  - version_five 4y ago
    The chatbot tools are unlikely to be faang material. I'd guess more likely the targets now are places like Salesforce that probably can find a way to claim value from adding some text generation.
- koito17 4y ago
  Yup. As far as I can tell, OpenAI is pricing at steep losses at right now. It makes me wonder what businesses and individuals will do once OpenAI runs out of money to subsidize everybody and massively increases their prices.
  - ano88888 4y ago
    well, paid users are increasing exponentially. 10 million paid users are 200 million monthly revenue already and it is just one source of their revenue.
rvz 4y ago
Sure... They sure 'really' are AI companies and totally not just wrappers around OpenAI or ChatGPT's API and giving Microsoft and Google more ideas to absorb for free with another AI bubble waiting to pop as soon as OpenAI increases prices. /s Then we will see how unprofitable these so-called AI hype 'startups' are.
- broast 4y ago
  Or these ai wrapper companies will get a head start on building the software technology to use LLM's while others work on getting these types of models (from open source or competing offerings) to run directly on our phones
- ilrwbwrkhv 4y ago
  It is just ridiculous made up stuff. I cannot wait to see how much damage a bunch of these AI companies cause. For example, I checkout out one which I won't name which does predictive Youtube analytics where they predict how many views you might get based on thumbnails and they are off by multiple orders of magnitude. Like it's not even close.
jay_kyburz 4y ago
I hope one of them is using machine learning to recognize photos of melanoma. I was thinking in the shower this morning that would be a handy service.
- palata 4y ago
  The recent buzz is over generative AIs. That kind of medical research has existed for a long time, it's not just part of the buzz (or not directly, at least).
pavlov 4y ago
I wonder if this gold rush is leaving openings for building simple solutions that don't need AI and emphasize traditional UX instead of conversational UI. If you have some domain knowledge, maybe you should focus on ensuring your product does a narrowly defined 100% job at what's needed, and it may shine when you're pitching against AI-based competitors whose products look like magic 95% of the time but are wrong or unusable the remaining 5%. "We are not a black box that sends everything to someone else's giant API" can be a selling point as businesses become more aware of their cloud dependencies.
- kolinko 4y ago
  The basic simple (web) solutions were built over the last 10-20 years - of course not everything was invented yet, but if there is a need for such a product, there is a high chance the product was already built, and the niche filled. Perhaps a better question is what new niches now open up, on top of the AI solutions. Hugging Face is a nice example - it's mostly a python library, a github clone, and a discussion forum strapped together. "Old" tech used to serve a new and growing niche.
  - geraltofrivia 4y ago
    HuggingFace is also hosting models (CLI/code/API), have an extensive community driven dataset store, and has some great positive feedback loops. Great Library -> Large Community -> Large Coverage of Models & Datasets -> Larger Community -> Larger Coverage of Models & Datasets -> Revenue -> More Engineers -> Greater Library ... Not all the steps (->) here are trivial: Contrast HF to Explosion (great folks behind spacy).
    - banditelol 4y ago
      > Not all the steps (->) here are trivial: Contrast HF to Explosion (great folks behind spacy). Can you elaborate more on this? about contrasting HF to Explosion?
      - geraltofrivia 4y ago
Mizoguchi 4y ago
Thin wrappers around ChatGPT may be ok to start. There's no point on creating your own GPT before validating your idea and getting some traction. Amazon didn't build their logistics and freight division first to then start selling books. Once your product or service is off the ground and more resources are available then you can start moving away from OpenAI. This could be much simpler that you think, particularly if you designed your systems with this in mind and considering future costs of training LLMs will be a fraction of what they are today.
marban 4y ago
Always thought that YC was meant to pick the Crème de la Crème of ideas. Different times indeed.
- whiplash451 4y ago
  They are still picking it up. It just happens to be sitting in the middle of a lot of noise.
truthsayer123 4y ago
So over 40% of YC batch are working on making API calls to OpenAI?
- whiplash451 4y ago
  Or building cheaper alternative to OpenAI.
  - truthsayer123 4y ago
    Do you know if it's actually true or did you make that up?
  - sottol 4y ago
    Not as far as I can tell? Any examples? Also, YC seed isn't gonna get you very far when compute alone for training an LLM is in the millions currently.
Krisjohn 4y ago
Is there still gold to be had, or are they just enriching the shovel-sellers?

63 points by charlierguo 4y ago

The companies that stuck with paper and didn't adopt computers, the web and mobile went out of business. The same will happen in the near future with the companies that are unwilling or unable to adopt AI/ML in their business.

palata 4y ago
Survivorship bias. You select technologies that won, and assume your new technology will be the same. What about all those technologies that just died? Companies that did not embrace Minitel, blockchain or the metaverse seem just fine right now.
bilqis 4y ago
Typical grifter take, using analogies, which are always a bad framework to analyse and predict events and trends based on the history. A few years ago others grifters were saying the same about blockchain and NFTs, a year ago same thing about metaverse/decentraland.
- borissk 4y ago
  Do tell - how do you analyse and predict events and trends NOT based on the history.
  - palata 4y ago
    A first good step IMO is to accept that we suck at predicting.
    - borissk 4y ago
      We are great at predicting. The number one task of the huge human neocortex is predicting the future. If you look at the past HN comments there were users who predicted the raise of OpenAI, the war in Ukraine and so on.
      - palata 4y ago
        Did you forget a "/s", or do you actually believe that?
        SideburnsOfDoom 4y ago
        No, but they are playing word games. yes, the human brain is a great engine designed for predicting future events and working with that e.g. how to catch a ball that is coming towards you and will arrive shortly in the future, how to throw an object and hit a target, how to pick up a heavy object and move it to where you need it, etc. But this is not at all the same kind of "predicting future events" as accurately saying where e.g. the CryptoCurrency or VR industry will be in a decade or so.
      - mym1990 4y ago
        Predicting isn’t some black and white skill. The neocortex doing calculations in the background to predict how heavy a cup of coffee might be or how hot the stove is…is very different from consciously predicting geopolitical events which are based on countless other variables. In a world of 8 billion people of course you will see some get things right, and then conveniently point it out…but most people’s predictions are wrong, and no one pays attention to those results.
  - SideburnsOfDoom 4y ago
    I'm puzzled by this take: surely the recent history - the cryptocurrency bubble - the hot air and lack of substance, the extractive scams, is useful lens for current events? But it's not everything, we can't be entirely "based on it" deterministically - as we do indeed "suck at predicting".
  - mym1990 4y ago
    That’s easy, if you have a framework of knowledge around things like physics or chemistry or biology(add any domain here), one could predict outcomes of hypothetical use cases with considerable accuracy. Unless you lump all domain knowledge into “history” because it has to have happened in the past…
    - Idiot_in_Vain 4y ago
      Do you think chemistry or biology knowedge was given to humans by the gods or the ancient aliens and not based on a history of practical knowedge, experiments and research?
      - mym1990 4y ago
        Thats why I added the last sentence. But for the purpose of debate, it is reasonable to distinguish domains like physics from history. It is one thing to study how something was discovered, and another to understand the parameters around an event so that you can predict what might happen.
  - bilqis 4y ago
    My point was, that saying “historical event X is somehow similar to ongoing event Y, thus we can reason and predict outcomes of Y based on X” is a false methodology, appealing, but false.
- nikanj 4y ago
  Remember when people paid good money for Second Life real estate?
- SideburnsOfDoom 4y ago
  Exactly. What was the corresponding percentage of Blockchain and/or Cryptocurrency start-ups say, 2 years ago?
Beaver117 4y ago
Except these companies are just thin wrappers over OpenAI APIs
- borissk 4y ago
  Today they may be thin wrappers over OpenAI APIs, but if they start bringing in big money they can hire data scientists and build their own models.
  - rvz 4y ago
    That is a big "if", which doesn't cut it with startups looking for VCs, given 95% of these ML startups depend on OpenAI's APIs. What I am looking for is 'even if'. So even if they make money AND hire 'data scientists', there is a higher chance that they will run out of money quicker since they are competing against Google, Microsoft and OpenAI which they all can afford to run everything at a loss for free and raise their prices which eats these startup's revenue. Either way, they are locked into these APIs and they aren't focused on profitability to even begin building their own models at this stage. If they are not profitable by the time the price increases come in, they are going to be struggling for survival.
dddw 4y ago
Frakkin' toasters everywhere!
kromem 4y ago
Yes, but there's a massive difference between a startup focused on a broader business issue incorporating AI into its plan on tackling that issue and a company focused on AI as the primary product. The former are going to mostly be successful, particularly in the place of companies that don't adapt, but the latter are mostly going to be obsolete by the time they'll be several years into it. AI as a product is a TERRIBLE play. But AI as part of a solution is going to be necessary for nearly every business vertical.
onion2k 4y ago
Some businesses absolutely have adopted IT with vigour and massively improved the way they work, but many haven't. They've just 'ported' their old paper processes to computers without learning or improving anything. Sending emails is faster than using paper memos but if you're just sending memos by email nothing is actually better. Using an Excel spreadsheet with no macros, formatting, or functions doesn't really give you much that double entry bookkeeping in a real book wouldn't already have given you. Plenty of people use Word with everything 'annoying' switched off, and they could do the same job on a typewriter. Similarly, lots of companies spent thousands of dollars on a website or a mobile app that no one looks at because they're a small town landscaping business that really doesn't need those things. The same goes for AI. There are definitely use cases where it will improve how we do things. For some businesses that will make them more profitable and more competitive. A lot of businesses won't adopt it for decades. Some will adopt it when it brings them no real benefits. AI is a tool. People will often use it poorly, just like every other tool.
choeger 4y ago
I have yet to encounter a single valuable (in terms of hard currency) use case for GPT besides automating publishing. Yes, you can generate text automatically now in relatively high quality. But that's a business that was struggling before GPT came along and I don't think we actually need that much automatically generated bullshit content. Yes, a language model will help authors to write their articles/books quicker, but that's not a significant fraction of their time spent, hopefully. I think language models will become a commodity, like a spell checker is today. They might open up the market for translations into more exotic languages (but I don't know how the quality is for such languages and by definition, the market ain't big) and I think every author will use one as part of their workflow (think of technical writers using a language model for the textual parts). Computer games might profit massively in terms of immersion. No, I don't think that new applications will become that important with language models. I think it's the other way around. There are so many applications that we simply don't need any longer. Language models and image recognition models together will take over the user interface industry. Why bother with web design and user experience if you can just throw the chatbot at your client? That's going to become the real value here. If I was a junior web developer, I would begin to seek out new programming languages and stacks.

wallaBBB 4y ago

This is not going to end well.

warent 4y ago
In what way?
- onion2k 4y ago
  One possible way it could end badly would be if OpenAI ramped it's pricing up to the point where it takes away most of those company's chances of ever being profitable. I'm not saying it's going to happen, and I hope it doesn't, but when I did a little analysis for an idea I have it was absolutely the biggest threat I could see. OpenAI has the ultimate vendor lock-in - there literally isn't an alternative right now. (This is a giant opportunity if you're a billionaire investor. Providing an AI backend is probably the next cloud infrastructure. OpenAI is going to be the AWS of that industry, but there's room for an Azure or a GCP too.)
  - Mizoguchi 4y ago
    There's really no point in worrying about OpenAI shutting down their API or pricing you out. It may happen but very unlikely that it will be suddenly and soon. Yes, keep it in the back of your mind and design your systems in such way that when the time comes replacing the GPT doesn't require rewriting your entire codebase. Your immediate #1 priority in such competitive market should be to execute your idea and get paying customers, then you deal with breaking off OpenAI.
  - whiplash451 4y ago
    OpenAI’s moat is not _that_ big in the grand scheme of things. You can train GPT2 in a couple weeks on a dozen GPUs. Sure, it’s a lot worse than GPT4, but not insanely behind and the community is fast. Same for RLHF. It is for sure fine engineering, but not unreachable for a motivated and competent bunch.
- wallaBBB 4y ago

derelicta 4y ago

oh yeah I've noticed this trend on indiehacker. Lotsa lotsa new products being advertised as AI-powered and whatnot. So it doesnt suprise me a lot.

fortissimohn 4y ago

Surely this technology will be used for the benefit of humanity. By 2030 we will be working 15 hour weeks and all people will have homes!

tuyguntn 4y ago
(for the sake of critical thinking) or rich get richer, poor get poorer and will not have access to AI tools, society will collapse, more jobs will be low paying, most on demand skills would be everything done by physical activity
- admissionsguy 4y ago
  > society will collapse, The great promise of robotics and AI is the possibility of decoupling societal stability from the vagaries of the poor.
  - borissk 4y ago
    How? By having armies of semi-intelligent armed drones, that will kill the poor if they decide to riot?
    - kolinko 4y ago
      By bringing wealth so cheaply that it can be given away. We already have many things that are given away for free, whereas a 100 years ago they were considered luxury (books, pens, papers, access to knowledge in general, and in Europe - public transport and healthcare, plus to some extent food, heat and electricity).
      - whiplash451 4y ago
        Then that’s not wealth anymore, because wealth is relative.
      - Idiot_in_Vain 4y ago
        Would the wealthy and the powerfull be happy for everyone to become wealthy?

raverbashing 4y ago

How many of those are just a minor prompt + API call to OpenAI?

yieldcrv 4y ago
Look, you’re not going to get investor money to train an LLM and scrape a ton of data sources at this stage Is what it is
mirekrusin 4y ago
Probably more than 90%?
onion2k 4y ago
The challenge of building, running, and growing a startup is rarely building some novel algorithm or complex machine. You can reduce almost all SaaS businesses down to "It's just a UI and a database!" if you want to, and many developers do. That only shows how easy it is to over-simplify things. My first startup was literally just some Javascript and a Postgres database. It didn't even call any external services, let alone one that's truly on the bleeding edge of computer science. I still thought it was a viable business (wrongly, it turned out, but still... I had a dream!).
- raverbashing 4y ago
  I find it funny some people being dismissive of what they built to defend the AI startups "Some JavaScript and a DB" is a web service. It does stuff. Sure if it's small or a basic crud then the moat is small But if I can replace the value proposition of a startup with "open chatgpt and type a short prompt" then there's not much value in it right. Maybe if there's value in the surrounding services.
  - onion2k 4y ago
    If the business is literally just sending data to GPT and outputting the response then sure, but I've not seen any businesses trying that yet. Pretty much everything I've seen to date is "A CRUD app with GPT to fill in blank data". That's automatically more valuable than just a CRUD app.

maz1b 4y ago

Surprisingly, there was only one (!) education startup in this batch. Doesn't bode well for us as we're applying in this Summer 23 cycle albeit with customers, revenue, PMF, etc.

palata 4y ago
[flagged]
- UncleEntity 4y ago
  “AI education” if you really want to get the money. Somebody has to teach the AI how to do things.
dang 4y ago
Don't worry about that at all! It will help you stand out.

danielmarkbruce 4y ago

Not surprising. You can sit down and bust out something pretty cool in an afternoon using the infrastructure currently in place (openai, huggingface etc). Makes more sense than being the 57th SaaS web app in any given space.

Beaver117 4y ago
If the bar for entry is that low how will these companies survive?
- palata 4y ago
  The probably mostly won't. But they don't care now. Now they want VC money. That's how startups work.
- nikanj 4y ago
  Every investment fund is under strict orders to plow money into AI, because it's going to change the world
- surgical_fire 4y ago
  They only need to survive for a long enough time until they find an exit. Either through acquisition or IPO. It doesn't need to be sustainable nor particularly innovative.
  - lma21 4y ago
    It's going to be tough to IPO when you're not sustainable or even profitable these days.
- kolinko 4y ago
  That's where the stat that 99% startups fail comes from. One or few winners, a hundred losers - the whole startup ecosystem works like this :)
- danielmarkbruce 4y ago
  The bar to making a social network weren't high from a technology perspective. Same for a lot of enterprise software.

ookblah 4y ago

feel kind of sorry (or excited?) for this batch lol. i have trouble believing that many companies were already working on AI/ML stuff. basically huge land/cash grab right now. surely at this level there has to be a lot of overlap... it really feels like just throwing everyone into the pit and having them duke it out and see who wins, loser be damned. i suppose that's the VC play in general, just a bit jarring to see it front and center.

alfiedotwtf 4y ago

"We're So Lo Mo... I mean So Mo Lo"

cornholio 4y ago

And most of those "companies" are shims over GPT. Zero moat, fake it till you make it mentality, gather investor money and maybe we'll roll our own in good time. Don't know how I feel about that, it quacks like a bubble but there is massive value that can be delivered.

whiplash451 4y ago
Sure. But as a VC you will look like a schmuck if you don’t invest in at least some AI startup.
nikanj 4y ago
It's the model familiar from previous bubble rounds: 1) Get funding 2) Put a lot of effort into appearances and presence 3) Get acquihired by Google / Microsoft / et al.
- whiplash451 4y ago
  Except step 3 is going to be a lot harder these days. You’d better have a strong story to sell something to Satya.
- version_five 4y ago
  The chatbot tools are unlikely to be faang material. I'd guess more likely the targets now are places like Salesforce that probably can find a way to claim value from adding some text generation.
koito17 4y ago
Yup. As far as I can tell, OpenAI is pricing at steep losses at right now. It makes me wonder what businesses and individuals will do once OpenAI runs out of money to subsidize everybody and massively increases their prices.
- ano88888 4y ago
  well, paid users are increasing exponentially. 10 million paid users are 200 million monthly revenue already and it is just one source of their revenue.

rvz 4y ago

Sure... They sure 'really' are AI companies and totally not just wrappers around OpenAI or ChatGPT's API and giving Microsoft and Google more ideas to absorb for free with another AI bubble waiting to pop as soon as OpenAI increases prices. /s Then we will see how unprofitable these so-called AI hype 'startups' are.

broast 4y ago
Or these ai wrapper companies will get a head start on building the software technology to use LLM's while others work on getting these types of models (from open source or competing offerings) to run directly on our phones
ilrwbwrkhv 4y ago
It is just ridiculous made up stuff. I cannot wait to see how much damage a bunch of these AI companies cause. For example, I checkout out one which I won't name which does predictive Youtube analytics where they predict how many views you might get based on thumbnails and they are off by multiple orders of magnitude. Like it's not even close.

jay_kyburz 4y ago

I hope one of them is using machine learning to recognize photos of melanoma. I was thinking in the shower this morning that would be a handy service.

palata 4y ago
The recent buzz is over generative AIs. That kind of medical research has existed for a long time, it's not just part of the buzz (or not directly, at least).

pavlov 4y ago

I wonder if this gold rush is leaving openings for building simple solutions that don't need AI and emphasize traditional UX instead of conversational UI. If you have some domain knowledge, maybe you should focus on ensuring your product does a narrowly defined 100% job at what's needed, and it may shine when you're pitching against AI-based competitors whose products look like magic 95% of the time but are wrong or unusable the remaining 5%. "We are not a black box that sends everything to someone else's giant API" can be a selling point as businesses become more aware of their cloud dependencies.

kolinko 4y ago
The basic simple (web) solutions were built over the last 10-20 years - of course not everything was invented yet, but if there is a need for such a product, there is a high chance the product was already built, and the niche filled. Perhaps a better question is what new niches now open up, on top of the AI solutions. Hugging Face is a nice example - it's mostly a python library, a github clone, and a discussion forum strapped together. "Old" tech used to serve a new and growing niche.
- geraltofrivia 4y ago
  HuggingFace is also hosting models (CLI/code/API), have an extensive community driven dataset store, and has some great positive feedback loops. Great Library -> Large Community -> Large Coverage of Models & Datasets -> Larger Community -> Larger Coverage of Models & Datasets -> Revenue -> More Engineers -> Greater Library ... Not all the steps (->) here are trivial: Contrast HF to Explosion (great folks behind spacy).
  - banditelol 4y ago
    > Not all the steps (->) here are trivial: Contrast HF to Explosion (great folks behind spacy). Can you elaborate more on this? about contrasting HF to Explosion?
    - geraltofrivia 4y ago

Mizoguchi 4y ago

Thin wrappers around ChatGPT may be ok to start. There's no point on creating your own GPT before validating your idea and getting some traction. Amazon didn't build their logistics and freight division first to then start selling books. Once your product or service is off the ground and more resources are available then you can start moving away from OpenAI. This could be much simpler that you think, particularly if you designed your systems with this in mind and considering future costs of training LLMs will be a fraction of what they are today.

marban 4y ago

Always thought that YC was meant to pick the Crème de la Crème of ideas. Different times indeed.

whiplash451 4y ago
They are still picking it up. It just happens to be sitting in the middle of a lot of noise.

truthsayer123 4y ago

So over 40% of YC batch are working on making API calls to OpenAI?

whiplash451 4y ago
Or building cheaper alternative to OpenAI.
- truthsayer123 4y ago
  Do you know if it's actually true or did you make that up?
- sottol 4y ago
  Not as far as I can tell? Any examples? Also, YC seed isn't gonna get you very far when compute alone for training an LLM is in the millions currently.

Krisjohn 4y ago

Is there still gold to be had, or are they just enriching the shovel-sellers?