Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
AI

GPT-4-Generated Pitches Are 3x More Likely To Secure Funding Than Human Ones (zdnet.com) 41

Posted by BeauHD from the one-step-closer-to-funding-Skynet dept.
An anonymous reader quotes a report from ZDNet: Clarify Capital, a small business lender, asked 250 investors and 250 business owners to rate a set of human-created and GPT-4-generated pitch decks without letting the participants know that AI was involved. To make matters more interesting, the human-generated pitches were successful ones that had secured funding in the past. The results showed that GPT-4 pitches were overall more effective than those made by humans. The AI-generated pitches beat out human ones in quality, key element description, and problem description.

According to the survey, the investors and business owners were three times more likely to invest after reading the GPT-4 deck than the human one, and they found the AI-generated decks twice as convincing. Furthermore, one in five of those professionals said that they would invest $10,000 more in the AI-generated pitches. The survey also tested the effectiveness of the decks across different industries, including finance, marketing, and investment. Unsurprisingly, across all of these industries, the GPT-4 decks were more successful in securing investments. The survey doesn't disclose which GPT-4-based AI chatbot the survey is using.
This discussion has been archived. No new comments can be posted.

GPT-4-Generated Pitches Are 3x More Likely To Secure Funding Than Human Ones More

GPT-4-Generated Pitches Are 3x More Likely To Secure Funding Than Human Ones

Comments Filter:

  • According to a random hunch, 2/3 of the investors asked, used GPT to evaluate the pitches.

    But seriously, they are looking for something unique. Any change in language, for better or worse, will make a pitch stand out. It probably had a lot less to do with the content itself than the perception that the writer is "different" somehow.

    • Is ChatGPT unique because it doesn't make grammar and spelling mistakes, unlike hu-mans?

      • It does have a flair for getting everything right, aside from factual accuracy and hallucinating information. I wish the average human could generate text as well as that thing can.

    • Re: (Score:2)

      by Calydor ( 739835 )

      Possibly it's also related to the certainty with which ChatGPT makes its claims. There is never any feeling of doubt in its mind, no feeling of not over-promising, no ... sense of reality, if you will. It's told to generate a winning script, and it will do so - whether able to follow through on it or not.

      • This overly confident, my product is glorious attitude in sales people is the one thing that stops me from buying. Always happy when they bring a technical guy along. First task? Try to get our sales people in a separate room with their sales people. Then just talk to the technical guy.
        Not easy since often the sales guy does want to keep everything under control on his terms.
        How long would it take the bad ones to start controlling GPT's output to tweak it to their vision?

    • Re:The Twist (Score:5, Insightful)

      by Joce640k ( 829181 ) on Saturday June 17, 2023 @09:25AM (#63610272) Homepage

      All this proves is that "investors" have no idea how to evaluate the real viability of a project, they just look at how shiny the presentation is.

      • Is capitalism's claim of efficiency just an AI-like con job?

        • Capitalism will generally decide much more quickly if something is a loser thus minimizing investment. Or rather, it used to before the rise of stakeholder capitalism. Incidentally, this combined with ESG has caused Disney's incipient downfall.

      • It also proves that pretending to invest is very different from investing.

  • a milestone (Score:5, Interesting)

    by ganv ( 881057 ) on Saturday June 17, 2023 @08:26AM (#63610134)
    Mark 2022-2023 as the years this happened. A large section of human activities that has used generation of effective marketing materials as metrics for excellence and allocation of resources has been destabilized. A rational path forward is to recognize that comparing marketing materials was always a bad way to allocate resources. But we are so deeply caught up in the supremacy of marketing rhetoric and images and video that it is very hard to rebuild systems that evaluate excellence on the basis of actual production of quality goods and services.

  • Making facts and cost/benefit analysis up works! (Score:3, Insightful)

    by marcobat ( 1178909 ) on Saturday June 17, 2023 @08:34AM (#63610152)
    If human had made up their facts and their cost analysis as much as the GPT engine did, they would have had just as much success, investors are well known to be gullible.

  • Those who can, do not manage investment funds (Score:4, Insightful)

    by dfghjk ( 711126 ) on Saturday June 17, 2023 @08:43AM (#63610176)

    People who see through bullshit work in technical areas; investors are not smarter, just incredibly entitled. Dipshits with money, that's Silicon Valley.

    This is the one and only reason Elon Musk is successful, he knows how to manipulate morons managing money.

    • Re: (Score:2)

      by sjames ( 1099 )

      Elon's advantage is even thinner. As a moron with money (thanks to his Dad), morons managing money are naturally predisposed to listen to him. Had he been born middle class, he would instead be that barista that keeps fucking your order up. The rest of it is just a matter of having enough money to keep trying until something pays off.

      Honestly, I believe our economy, technology, and even our culture would advance better if we rounded up the morons in finance and dropped them on an island somewhere. It's time

    • Re: (Score:2)

      by gweihir ( 88907 )

      That is pretty accurate.

  • Because writing is hard (Score:4, Insightful)

    by iAmWaySmarterThanYou ( 10095012 ) on Saturday June 17, 2023 @08:45AM (#63610184)

    Clear, effective writing is a serious skill requiring years of experience and feedback, as well as editing, etc etc etc.

    The typical tech founder has essentially near zero quality writing experience.

    So, yeah, the pitch made by a computer basically rehashing previous successful pitches is not surprisingly going to be better than some developers who have trouble forming coherent sentences in English.

    • But us humans are still better at lying! You left that part out.

      • Lol, I was about to agree but no, chatgpt has any number of documented hallucinations. It's really good at lying.

      • ChatGPT doesn't have a conscious, so it probably does a better job creating pitches telling investors what they want to hear that isn't based on economic reality or realistic time constraints. It would probably make a good tech CEO as well, as it seems to have it's own built-in reality distortion field regarding technical details.

        • Re: (Score:2)

          by g01d4 ( 888748 )

          It would probably make a good tech CEO as well,

          Exactly. All the known 'faults' of current LLMs, such as hallucinations, are bonuses in a world of unicorns.

        • Very few successful startup CEOs/founders have a conscious or sense of ethics. If you read the stuff put out by venture guys and people in the business of picking new founders to fund, etc, those are considered positive traits in that field. It is well accepted that the typical CEO is a narcissist and sociopath.

  • What you're saying here is that ChatGPT is very good at being mellifuous.

  • Not impressive (Score:3)

    by RonVNX ( 55322 ) on Saturday June 17, 2023 @09:00AM (#63610208)

    Impressing the VC class is like impressing people who drool on themselves.

  • To me while this shows that AI is still advancing on the other hand it kinda confirms that VC funders aren't these wise sages who have the gift to pluck the real geniuses out of obscurity but still just people reacting to their subjective whims.

    It would be interesting to see the success rates in a few years of these companies. Did the AI ors succeed? Isn't that the metric we are really after?

    This really makes what we all think happens does happen. Put enough razzle dazzle and buzzwords to get funded. Aft

  • I'd hazard a guess that gpt4 has a snapshot of the trending buzzwords of the past 5 years. Once a critical mass of pitches are machine generated and these are used in newer models the entire thing will disappear up its own asshole.
  • Since we know that most "pitches" are full of "unrealities", does this mean ChatGPT is a better liar than a group of humans?

  • This is a biased antidote to support a sales pitch. Thanks for that slashdot.

  • Just like pop music (Score:3)

    by MpVpRb ( 1423381 ) on Saturday June 17, 2023 @11:43AM (#63610486)

    Chatbots are able to generate bland mediocrity that exactly conforms to expectations

    In pop music, expert teams of composers, arrangers, lyricists, producers, engineers, trendmongers, fashionistas, marketoids, choreographers and other assorted specialists create artless, mercenary, industrial product, expertly crafted to sound familiar but just the tiniest bit different enough to not be exact copies of existing hits

    This is exactly the same thing that chatbots do

  • Investors were given pitch decks written out in longhand, and also typed on a typewriter. They preferred the typed ones 2:1.

    In other words, everyone will use chatgpt to write pitch decks and we can all stop talking about it like it's some immense revolution.

  • the only relevant info from the methodology:
    "...we provided GPT-4 with various prompts for each industry to generate pitches."

    Various prompts - so it's possible that the chatbot entirely created the pitch, or it might have done some editing/polishing of the "prompt" - or anything in between.
    This bullshit is not worth reading, the less commenting...

    Remember the firm that produced this fake study: Clarify Capital. And avoid them.

  • What this really means ... (Score:5, Insightful)

    by Big Bipper ( 1120937 ) on Saturday June 17, 2023 @12:22PM (#63610594)
    What this really means, is that ChatGPT has learned to lie as convincingly as a good used car salesman. If we don't know that we are talking to something with the equivalent morals, we don't automatically set our bullshit detectors to max.

  • The human tries to make money by telling the truth.

    The bot tries to get investors by saying whatever they want to hear.

    Not surprising that the bot gets funded more. What would be surprising is if the bot makes more money.

  • It has been asked as a joke, but the question seems interesting to me: how does an AI rate business models made by AI versus humans? Is it possible to tell them apart? Humans seems able to unconsciously do so, since they prefer the GPT-4 productions.

  • If you want a sales pitch that invents a new reality and leaves out any weasel words, then this is a job for a machine and not a human. A machine will not hesitate to lie to you. And I assume we're not far off to a world when a machine will not hesitate to kill.

Slashdot Top Deals

It's a naive, domestic operating system without any breeding, but I think you'll be amused by its presumption.

Close