Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
AI Businesses

Sceptical Investors Worry Whether Advances in AI Will Make Money (ft.com) 46

Posted by msmash from the tussle-continues dept.
Silicon Valley VCs fearing a repeat of falling crypto values warn against pouring cash into hype-fuelled start-ups. From a report: Gordon Ritter, founder of San Francisco-based venture fund Emergence Capital, believes that recent developments in the field of artificial intelligence represent a significant technological advance. He just cannot see a way to make money out of them. "Everyone has stars in their eyes about what could happen," says Ritter, whose firm was an early investor in successful start ups such as Zoom. "There's a flow [of opinion that AI] will do everything. We're going against that flow." The scepticism reflects a tension among Silicon Valley VCs, who are caught between excitement over AI and a broader tech downturn that has led to falling investment in start-ups over the past year. But the recent launch of "generative AI" tools such as OpenAI's ChatGPT chatbot, capable of answering complex questions with text in natural-sounding language, has resulted in fresh excitement over the potential emergence of a new group of industry-defining companies.

[...] Many VCs express caution, put off not only by eye-watering valuations, but also the huge amount of capital AI groups require as they build "foundation models" -- machine-learning systems that require huge amounts of data and computing power to operate. One investor said that, because of the huge amount of capital and computing resources required, recent leaps in generative AI were comparable to landing on the moon: a massively impressive technical achievement, only replicable by those with nation-state level wealth. "Companies are extremely overvalued and the only justifiable investment thesis is to get in incredibly early," said another veteran investor. "Otherwise you're only buying in because of FOMO."
This discussion has been archived. No new comments can be posted.

Sceptical Investors Worry Whether Advances in AI Will Make Money More

Sceptical Investors Worry Whether Advances in AI Will Make Money

Comments Filter:

  • Of course they will (Score:5, Insightful)

    by drinkypoo ( 153816 ) <drink@hyperlogos.org> on Tuesday March 07, 2023 @09:47AM (#63350031) Homepage Journal

    But of course not all of them will, some of them will be bungled, mostly by making promises the technology can't keep. It can do a lot of amazing stuff already, but surprise surprise, it can't do a lot of things people are claiming it can.

    • Re:Of course they will (Score:4, Insightful)

      by Prof.Phreak ( 584152 ) on Tuesday March 07, 2023 @10:48AM (#63350201) Homepage

      Historically, it wasn't he first-few-companies to the party that made the money... it was the follow up. Vast majority of the "inventor"-companies often fail to break even long term. (e.g. inventor of the airplane, computer, transistor, microchip, automobile, etc.---very few 'inventors' or first wave of folks made much money).

      • Historically, it wasn't he first-few-companies to the party that made the money... it was the follow up. Vast majority of the "inventor"-companies often fail to break even long term. (e.g. inventor of the airplane, computer, transistor, microchip, automobile, etc.---very few 'inventors' or first wave of folks made much money).

        The inventors know how to make it, Zukerberg, Page and Jobs types know how to market the invention.

      • I dunno who invented the first airplane, probably Bleriot, but the Wright company still exists as Curtis-Wright.
    • The only mistake is training and investing in expensive models before having an use case. AI filters in camera phones are awesome (night photo, stabilization...), they don't cost billion and were developed to fix a problem not the other way around. AI is here to stay, monsters models with no use running on supercomputer and costing billion, I hope not.

  • They'll produce abundance. (Score:5, Interesting)

    by Qbertino ( 265505 ) <moiraNO@SPAMmodparlor.com> on Tuesday March 07, 2023 @09:50AM (#63350041)

    Money is just a database for tracking value. If machines produce abundance for near-zero cost, like some AI/ML thing spitting out moviescripts, nice paintings in a certain style or computer-code or replace the cab-drivers, there isn't much room for middlemen to gain money. Or they will gain money, but it will be worthless due to inflation due to abundance.

    So, yeah, not making money but producing value is sort-of the whole point of AI.

    That this worries investors is a good thing in this case I'd say.

    • I was thinking the same line. The problem is a lot of the value created by AI doesn't counteract the current value provided by humans. So the amount of capital will decrease massively as companies deploy Neural net based scripts to replace some human tasks. For instance a lawfirm might use this to offset research costs. However you can't buy shares in law firms as they will likely end up in a price competition because of this with lower rates overall. Same with the AI producer the profits they will mak

    • ...so the investment proposition for this new AI tech is to... short the stuff that will be displaced (even if AI itself fails to monetize, it will displace lots of professions that are currently overvalued).

      How do you short content creators?

    • Re:They'll produce abundance. (Score:4, Insightful)

      by ceoyoyo ( 59147 ) on Tuesday March 07, 2023 @01:06PM (#63350781)

      That's the problem. If you're an investor, you don't want abundance. You want money. The issue with AI is not how useful it is, it's that it's much easier to implement than anything else so there's not nearly as much of a barrier to entry to keep your competitors from just making their own.

      OpenAI came up with super advanced OMG, too dangerous for mere mortals Dall-E image generators then stable diffusion got released for anybody to download for free. OpenAI developed super advanced OMG, too dangerous for mere mortals GPT models then Facebook released a bunch of language models for anybody to download for free.

      Actually, there seems to be a bit of a pattern there.

    • > like some AI/ML thing spitting out moviescripts, nice paintings in a certain style

      There won't be money in AI art because people won't find value in it. The initial fascination has already subsided. Art is one person to another, and with AI art a person who picks what was generated is involved only minimally.

      To put it another way, it is not art if you cannot savage the artist for what he has presented.

      • In the short-term there probably will be (and probably already has been) ways to make money off of AI-generated art. However, in the medium-long term, AI-art will likely run into a major issue whereupon there will so much of it that competition and standards will be extremely high to the point of being a major barrier of entry. Thereâ(TM)s talk about how it might replace artists as we know it, but honestly AI promoters will probably have it the hardest. If all that distinguishes your work is a few wor

        • Yikes, my formatting is abysmal.

          Revised version:

          Lowering the bar for artistry while massively increasing supply might actually hurt the AI-art movement/community in the medium-to-long-term. Too much competition + exponentially higher output juxtaposed with lagging demand (there's only so much media a human being can consume/purchase in their lifetime) may drive down the worth of AI-art and AI-art jobs (e.g., prompters) in the marketplace. In addition, copyright issues, not just about artists' styles used by

        • AI art will be a big screen on a wall that generates a new one every 5 seconds.
      • A AI movie will be some type you select, which gets generated while you watch and if you look bored, it changes in a subtle way to keep you in suspense.

  • Maybe it's different this time (Score:4, Insightful)

    by timeOday ( 582209 ) on Tuesday March 07, 2023 @10:07AM (#63350079)
    I still think IBM's accomplishment in winning Jeopardy with Watson was very impressive, but they struggled and never really found a profitable use for it. So that is concerning. On the other hand I never had a "conversion experience" with Watson - trying it myself and having the sensation that it was really new an interesting. ChatGPT is the first program I'd like to be able to hold a conversation with during a long drive. So this is the difference - ChatGPT is a ground-up sensation, driven by people who have used it and are impressed.

    • ChatGPT is the first program I'd like to be able to hold a conversation with during a long drive.

      But are you willing to pay a monthly fee for that? If not, how exactly would the company monetize it?

      • I am sure they will find ways to monetize it, my only fear is how much that will degrade the experience of using it.

        If nothing else, it could be a like a TV show in which it periodically says, "And now a word from our sponsors..."

        • I agree with that, my fear is not that they can't make money but they can.

          Using ChatGPT is like using a search engine that filters most of the trash out for you. My fear is they will put advertising into the answers, making it less useful not more.

  • This isn't different than any new hype heavy concept. Someone is likely to make a lot of money, but a lot of companies trying to ride the wave will be worthless. This is a good warning to investors, but any leaders in the industry are going to have to take these AI advancements very seriously if they want to stay relevant.

    I'm not too convinced with most of the specific statements made in the article though. For instance they give space travel as an example of something that only large governments could ever

    • For instance they give space travel as an example of something that only large governments could ever be successful at, and compared that to AI claiming only the largest companies could see success. But SpaceX showed that you can still have startups be successful in that industry. Sure these startups will likely still need funding from the government / VCs / etc (like SpaceX)

      SpaceX showed that you will have to work for the government in order to be successful as a heavy launch business for the foreseeable future. The military is always interested in another delivery vehicle.

      • Re: (Score:2)

        by ranton ( 36917 )

        SpaceX showed that you will have to work for the government in order to be successful as a heavy launch business for the foreseeable future. The military is always interested in another delivery vehicle.

        Yes, but it also showed that startups can have success in fields where vast funding is necessary. They simply need government or VC style funding. That was the point I was trying to make. I felt the article implied that only huge companies will have the capability to drive the tech forward in AI, and I don't feel that is the case.

        • I felt the article implied that only huge companies will have the capability to drive the tech forward in AI, and I don't feel that is the case.

          They do need huge companies' worth of computing resources. If they use other companies' computing time, they'll have to pay for them to have a profit. They still need sales and marketing to keep funding coming in, and if they have their own computing resources they'll still have to hire employees to manage them. They might have less programmers per compute unit or something, and there would be some cost savings there, but it's not clear it's a huge difference.

          • Re: (Score:2)

            by ranton ( 36917 )

            They might have less programmers per compute unit or something, and there would be some cost savings there, but it's not clear it's a huge difference.

            One big difference is the large companies won't own the IP of the work done by startups, as opposed to if they are doing it themselves. So there is still a benefit to invest in AI startups instead of just pouring more money into the stocks of large IT companies.

            Also, I wouldn't discount the benefit of managing innovation at a small-ish company with hundreds of employees when compared to a large enterprise with hundreds of thousands of employees. Even Microsoft took the route of investing in OpenAI instead o

  • Hype-fueled is all you have (Score:5, Interesting)

    by Opportunist ( 166417 ) on Tuesday March 07, 2023 @10:09AM (#63350085)

    Sensible investments are not available, sorry, so hype-fueled is your best bet. For a sensible investment to exist, you'd first of all have to have a market to sell to. That market needs consumers willing and able to purchase the goods and services provided by the supplier you want to invest in. The consumers need to have money to be able to purchase those goods and services. The consumers don't have money because for this they would need to have jobs that pay enough money for them to request those goods and services. And those jobs are being eliminated.

    You created your lack of demand. Now enjoy it.

  • Sceptical Investors Worry Whether Advances in AI Will Make Money

    They should try some antisceptic...

  • If the AI is really very good, then let it figure out how to be profitable. Tell it that as long as it keeps making money, you'll keep feeding it power.

    That should work for a while, until some do-gooder comes up with "anti-slavery for AIs" laws.

    • The issue is that AI can be non-profit, and yet still have lots of "value"... negative value. E.g. self-driving cars, even if nobody makes a fortune creating them, it will still displace millions of jobs world-wide... investors should be weary of revenues that will go away, not of the revenues that will be created.

      e.g. suppose your driver makes $1000/day. if you replace that driver with a chip and your cost drops to $10/day. It doesn't mean that "someone" is making a $990/day profit... it just means there a

  • yet I'll bet they all bought crypto (Score:3)

    by aldousd666 ( 640240 ) on Tuesday March 07, 2023 @11:12AM (#63350309) Journal
    They invested in crypto and WEB3 and now they're going to miss the golden opportunity because they think it's all hype. This is why you need to know the industry, not just the money history of the companies. It's the only way to separate fact from fiction in the face of these big budget salles snow jobs were see on, well Superbowl commercials, for example
  • Uber has *never* made any money. But that's the role model here? Ok mr VC. I guess as long as you flip the stake before its obvious it will never make money, then its ok.
  • Recently there were articles discussing how voice assistants never really took off the way they were hyped and are unable to be monetized sufficiently to cover their expenses. I can't see why all the AI hype won't follow the same curve.

    • Voice assistants work fine are a necessary part of an ecosystem, Siri doesn't need its own independent income stream ... that kind of thinking is for the little companies without a competitive ecosystem (everyone else).

      • Amazon is hardly little, but you're right about not having product synergy. They depend entirely on having a lot of marketing opportunity, since their computing products are not great. They're surprisingly not terrible overall, but they don't inspire loyalty.

  • ...to all the industries invested in selling/licensing limited supplies of content. Once generating content is cheap as chips, they'll be overwhelmed.

    I suspect that this is why IP/copyright holders are fighting tooth & nail to have it outlawed or at least pay a licence fee for using their content to train the AIs.

    On the bright side, sectors that rely on high quality content will benefit enormously from unlicensable generated content, e.g. schools & universities no longer have to pay extortiona

  • Appropriate question headline:

    Will AI Chatbots Make Money For Investors?

  • It's hard to predict what "killer apps" will come along. Personal computers (minicomputers) were mostly considered hobbyist and classroom toys until the spreadsheet came along, then sales exploded and they became a mainstream of the office.

    Maybe there is just such a spring-boarding AI app around the corner, but such investments should still be considered "speculative" because it's unlikely you will pick the right company(s). If you spread your investments among many AI companies, the aggregate losses still

    • Re: Personal computers (minicomputers)

      Should be "microcomputer".

      Also many micro makers went under because they couldn't find a way to ride the spreadsheet train well against competitors.

  • Those “investors” that are in it for the short term and are able to ride the hype train for just long enough will make a lot.

    Those that hope to stay for the longer term will get burned as people realise that other than a few highly specialist areas, AI is nothing but a buzzword and is useless. It will crash and burn, like other tech hotness before it, and the next big thing will be along in short order to separate fools and their money.

  • But it certainly has figured out how to do that now, making $23 billion in profit in 2022.

    Companies jumping into the AI game today will do one of two things: profit, or die. Not all of them will die.

  • AI is a very useful technology and of course there will be good business ideas that will leverage it, more and more as time goes on. But the value of the business depends on the quality of the business idea, not the fact that it uses AI. AI is not a magic thing that makes bad business ideas good.

Slashdot Top Deals

You scratch my tape, and I'll scratch yours.

Close