Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GPT-4 Outperforms Humans in Pitch Deck Effectiveness

clarifycapital.com

Post reply on HN
Print view View original

21–30 of 90 posts

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#21

Post by EA-3167 »

moglito wrote: post #17
dbfclark wrote: post #11

The key question here isn’t so much whether GPT-4 beats the actual human decks as much as what it had to fabricate to do so. The humans are probably constrained by things like “reality” and “what their business has done in it” while GPT-4 could make up anything it wanted. A fair comparison would be to humans given the same prompt and told to invent whatever facts they wanted.

Great point! Since they are not showing the decks, for all we know these GPT4 pitches may have been proposing they found a cure for cancer.

"In addition, my plan guarantees every child a free kitten or puppy, along with the aforementioned immortality."

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#22

Post by TechBro8615 »

ThomPete wrote: post #7

In my experience having pitched to VC's it's not that there is a perfect deck but many really bad ones. I.e. there is diminishing returns on a "perfect" pitch deck. You just want to be in the category of not sucking and the decks level of perfection is no longer important. Today I would even go as long as to say the pitch deck is much less important than a demo. You can in theory pitch in an email, the rest is just t…

The secret to raising money is getting the first offer. Everything falls into place after that, because investors are herd animals. Your pitch could go in one ear and out the other, but they'll suddenly start paying attention if you end it with "we've got an offer for $X at $Y cap, it's not from our ideal investor, and we'd prefer you lead the round, but we told them we'd get back to them by Wednesday and we just want to get back to work on building this thing."

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#24

Post by Ankaios »

dbfclark wrote: post #11

The key question here isn’t so much whether GPT-4 beats the actual human decks as much as what it had to fabricate to do so. The humans are probably constrained by things like “reality” and “what their business has done in it” while GPT-4 could make up anything it wanted. A fair comparison would be to humans given the same prompt and told to invent whatever facts they wanted.

The humans are probably constrained by things like “reality” and “what their business has done in it” while GPT-4 could make up anything it wanted.

You clearly know different founders than I do.

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#25

Post by usaar333 »

golergka wrote: post #20
usaar333 wrote: post #13

I find it weird to compare GPT-4 to not expert humans (which this did - they seemed to have just sampled pitch decks for funded companies). GPT-4 generally will beat non-experts. What you want to see is a comparison to say pitch deck consultants - it's also more useful to evaluate whether AI or a human should help you build one. This does suggest at minimum though that if you are a founder that GPT-4 will likely prov…

If a company was funded, doesn't it mean that pitch deck was successful?

Only on a binary scale. There's even among actually funded companies better or worse pitch decks.

You can often just get funding by having a good team matched to a VC that knows your industry. Pitch deck can be crap and other VCs might hard reject.

Another potential methodology flaw is that they don't seem to group decks by VC industry. I definitely altered my deck for different VCs, in ways which could make it a worse deck for the "average VC". GPT-4 might be better at setting context that industry expert VCs don't need - meaning this isn't even comparing like and like.

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#27

Post by paulddraper »

dbfclark wrote: post #11

The key question here isn’t so much whether GPT-4 beats the actual human decks as much as what it had to fabricate to do so. The humans are probably constrained by things like “reality” and “what their business has done in it” while GPT-4 could make up anything it wanted. A fair comparison would be to humans given the same prompt and told to invent whatever facts they wanted.

Right!?!

Ideally the most important thing in a pitch deck are the actual facts and substance, not the presentation and storytelling.

So what are the facts used?

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#28

Post by SamBam »

EA-3167 wrote: post #21
moglito wrote: post #17

Earlier quoted context omitted.

Great point! Since they are not showing the decks, for all we know these GPT4 pitches may have been proposing they found a cure for cancer.

"In addition, my plan guarantees every child a free kitten or puppy, along with the aforementioned immortality."

Sure, but a human could write that too. Surely the onus is on the investor to assess how realistic they think the founders' claims are.

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#29

Post by DotSauce »

I bet I could improve your prompts and make the pitch decks even better. Let me know if you would like a prompt audit. I also have a detailed prompt for content analysis and scoring. The reason being is because I do tests like this as well to compare GPT-4 basic prompts to prompts enhanced by the SmartGPT AI tool I created.

Re: GPT-4 Outperforms Humans in Pitch Deck Effectiveness

#30

Post by monnok »

moglito wrote: post #3
assbuttbuttass wrote: post #2

This says more about VC's than it does about AI

I'm in agreement regarding more . But I think it also says something about clear messaging and how many people struggle with that. It's actually consistent with YC advice. They say most pitches fail to even convey what the hell they are doing and what problem they are solving. I suspect a language model will get that part across rather well.

To be fair to humans, 90% of the workday is spent using bizarre internal corporate speak that has evolved for indirectness approaching meaninglessness.

It’s asking a lot to recognize a real transactional moment, codeswitch, and present effectively to the narrow goal. Even then, the narrow goal (maximize the chance of external funding) might be misaligned with the broad personal goal (don’t overcommit… on fucking anything).

A robot can just maximize on the funding result. It didn’t spend its whole career being lectured in circles about the next “one-corpco initiative” that will replace everybody’s SAP front-ends or whatever.

Post reply on HN