Hacker News Reader

Comment by typs | Hacker News Reader

ChatGPT is an extremely successful consumer product. Looking at Search traffic alone indicates they clearly have large brand value. Usage doesn't just depend on just benchmarks for most consumers.

View on HN

There's large brand value, and then there's $300B brand value.

Without a moat, it's hard to argue OpenAI's value is that high.

You can choose as an investor between:

a) Coca-Cola (~$300B, profitable, stable, stock-listed, well-known addictive brand, accepted business model with steady growth)

b) OpenAI (bleeding cash: $5B lost last year, tech that is not really novel anymore, weird corporate setup, competition that offers similar products).

It's not that easy to support the idea that OpenAI will be the winner.

I've been saying the same thing about Tesla for years, but here we are. The market can stay irrational blah blah blah

Well now that the Donald is running the US and is about to cut subsidies for electric vehicles, add tariffs to aluminum and steel, now maybe the markets may see the light, I am not holding my breath though.

Steel and aluminum tariffs will hit all car manufacturers simultaneously.

If all cars are more expensive, and all manufacturers have lower margins Tesla will have lower margins too. For the stock price the competition won't be other car manufacturers but other industries.

It would be a "subsiding tide lowers all boats" kind of thing.

I've been thinking, when I younger and living in the UK, all sorts of things (including some with the word "tax" in them) were denounced by the local right wing as "stealth taxes".

Trump is talking about tariffs as if they're a tax on non-Americans, but they're paid for by Americans who import stuff, which is basically all Americans given where your oil, aluminium, and steel come from.

Those tariffs are a stealth tax.

Cutting subsidies would weaken Tesla's competitors (according to Musk)

Tesla- being more vertically integrated- stands to benefit from disproportionately increased costs for its competitors

Just because costs are higher for its competition doesn't mean that people are going to be buying more Teslas. Last I checked, Tesla was already down 40-60% in sales at its major European markets. They're getting absolutely slaughtered by BYD in China.

I mean as long as people are still buying cars, Tesla becomes more cost competitive. But in the short run it's not for consumers as much as it is for investors.

You may be right that Tesla will struggle internationally and the car market as a whole will struggle in USA

That comparison also shows that a recognizable brand alone can easily be worth more than 100 billion; Coca-Cola also has competition that offers similar products and zero tech advantage, but they still managed to keep their number one spot for decades.

OpenAI is pre-revenue, they're a pure play.

It's not about how much you earn, it's about how much you're worth.

Marking “new tech paradigm isn’t about turning a profit/ the rules have changed” on my tech bubble bingo card.

For those not in the know: https://www.youtube.com/watch?v=BzAdXyPYKQo

the 'tres commas' era feels so quaint these days.

Is it already 1997 again?

Two bets with the same expected value can have wildly different variance.

The OpenAI bet is clearly that they build AGI and become the worlds first $10T company. At this valuation you only need to think it’s P > 0.1 for it to be a reasonable bet.

Coke is not a “swing for the fences” investment.

The choice is not between that. It's between which of the two businesses can deliver outsized growth & returns back to the investor in their time horizon.

Well, on the whole you can't because OpenAI isn't listed. If it was, retail investors would be piling in.

Huh? These are very different investment goals, you're not choosing between a or b

So invest all your money in Coca-Cola and maybe beat inflation by one or two percent.

Or you can gamble it all on a meme stock and maybe 100x your wealth.

Institutional investors are doing both.

OpenAI's real moat is all in their know-how about developing LLM architectures and training them. That stuff can be expected to be very hard to replicate, no matter how hard anyone tries. The best case for OpenAI is actually if the DeepSeek R1 improvements are genuine, and OpenAI can adopt them to make their upcoming models even more capable than they would be otherwise.

The best moat is that their technology is used by microsoft - who forced it on tons of paying subscribers.

Of course microsoft can switch to a better model, but a better model will not easilu build such a such a big subscriber base.

Now openai is powering tons of microsoft tools which are used by tons of people due to office integration.

> indicates they clearly have large brand value

BlackBerry, Nokia, Yahoo, and MySpace were popular at one point but couldn't keep up.

Yahoo is still website number 7 or 8 in the US.

Interesting. I would not have guessed that. https://www.similarweb.com/top-websites/united-states/

I don't know a good source for MAU numbers, but they would have to be very different with mobile app usage included.

I would guess that OpenAI (possibly as part of Microsoft or another company) will still be in the top 7 AI companies in a few years.

Conversions to paid users us only 5-6% though

https://beta.bnnbloomberg.ca/business/company-news/2024/10/2...

Does that number matter?

What’s the conversion rate on any single typical ad hosted by Google? Yet it all adds up to make a lot of money, for Google and (presumably) the advertisers.

Indeed it matters. A popular website is only financially valuable if the site owner runs ads on it, as Google does. Without ads, ChatGPT's consumer business model is not much different from WinRAR with its soft 60-day trial period. Sure, some % of its users upgraded to the full thing. Everybody else had humbler use cases covered by the free functionality (the 60-day expiry was famously not enforced).

Once zip/unzip technology became commodified (either by being baked into the OS or open-sourced as happened with 7zip), paid 3rd party tools were no longer needed and WinRAR's business disappeared.

As more advanced AI features get bundled into apps that users are already using like Google Search, FB, etc. there will be less incentive to have a separate ChatGPT subscription (even if it's slightly better, it won't matter to most people).

The one bundled to google search is so poor quality that it is degrading google experience...

How much of that community would jump ship on change of ownership? Not that Altman is any better than Musk but Musk has certainly created a more negative publicity about himself recently and thus more distrust or people of certain political stripes unwilling to use his products (scientific community probably leans that direction).

I've soured on Altman, but he's MUCH better than Musk. It's not even a competition.

they just need to find some laws to break or some poor people to exploit so they can be like other big SV startups.

Reply on news.ycombinator.com