[fairity]

DeepSeek just further reinforces the idea that there is a first-move disadvantage in developing AI models.

When someone can replicate your model for 5% of the cost in 2 years, I can only see 2 rational decisions:

1) Start focusing on cost efficiency today to reduce the advantage of the second mover (i.e. trade growth for profitability)

2) Figure out how to build a real competitive moat through one or more of the following: economies of scale, network effects, regulatory capture

On the second point, it seems to me like the only realistic strategy for companies like OpenAI is to turn themselves into a platform that benefits from direct network effects. Whether that's actually feasible is another question.

[aurareturn]

This is wrong. First mover advantage is strong. This is why OpenAI is much bigger than Mixtral despite what you said.

First mover advantage acquired and keeps subscribers.

No one really cares if you matched GPT4o one year later. OpenAI has had a full year to optimize the model, build tools around the model, and used the model to generate better data for their next generation foundational model.

[dplgk]

What is OpenAI's first-mover moat? I switched to Claude with absolutely no friction or moat-jumping.

[xxpor]

What is Google's first mover moat? I switched to Bing/DuckDuckGo with absolutely no friction or moat jumping.

Brands are incredibly powerful when talking about consumer goods.

[bpt3]

Google's moat was significantly better results than the competition for about 2 decades.

Your analogy is valid at this time, but proves the GP's point, not yours.

[fairity]

I think it's worth double clicking here. Why did Google have significantly better search results for a long time?

1) There was a data flywheel effect, wherein Google was able to improve search results by analyzing the vast amount of user activity on its site.

2) There were real economies of scale in managing the cost of data centers and servers

3) Their advertising business model benefited from network effects, wherein advertisers don't want to bother giving money to a search engine with a much smaller user base. This profitability funded R&D that competitors couldn't match.

There are probably more that I'm missing, but I think the primary takeaway is that Google's scale, in and of itself, led to a better product.

Can the same be said for OpenAI? I can't think of any strong economies of scale or network effects for them, but maybe I'm missing something. Put another way, how does OpenAI's product or business model get significantly better as more people use their service?

[nyrikki]

You are forgetting a bit, I worked in some of the large datacenters where both Google and Yahoo had cages.

1) Google copied the hotmail model of strapping commodity PC components to cheap boards and building software to deal with complexity.

2) Yahoo had a much larger cage, filled with very very expensive and large DEC machines, with one poor guy sitting in a desk in there almost full time rebooting the systems etc....I hope he has any hearing left today.

3) Just right before the .com crash, I was in a cage next to Google's racking dozens of brand new Netra T1s, which were pretty slow and expensive...that company I was working for died in the crash.

Look at Google's web page:

https://www.webdesignmuseum.org/gallery/google-1999

Compare that to Yahoo:

https://www.webdesignmuseum.org/gallery/yahoo-in-1999

Or the company they originaly tried to sell google to Excite:

https://www.webdesignmuseum.org/gallery/excite-2001

Google grew to be profitable because they controlled costs, invested in software vs service contracts and enterprise gear, had a simple non-intrusive text based ad model etc...

Most of what you mention above was well after that model focused on users and thrift allowed them to scale and is survivorship bias. Internal incentives that directed capitol expenditures to meet the mission vs protect peoples back was absolutely a related to their survival.

Even though it was a metasearch, my personal preference was SavvySearch until it was bought and killed or what ever that story way.

OpenAI is far more like Yahoo than Google.

[rayval]

In theory, the more people use the product, the more OpenAI knows what they are asking about and what they do after the first result, the better it can align its model to deliver better results.

A similar dynamic occurred in the early days of search engines.

[aurareturn]

They have more data on what people want from models?

Their SOTA models can generate better synthetic data for the next training run - leading to a flywheel effect?

[eikenberry]

Google wasn't the first mover in search. They were at least second if not third.

[talldayo]

> What is Google's first mover moat?

AdSense

[xxpor]

But _why_ did AdSense work? They had to bootstrap with eyeballs.

Claude has effectively no eyeballs. API calls != eyeballs.

[tucnak]

It's like people forget Google is an ad company

[ttoinou]

But most of the money to be made in AI is B2B, no ? Not direct consumer products like ChatGPT being used by the public

[moralestapia]

*sigh*

This broken record again.

Just observe reality. OpenAI is leading, by far.

All these "OpenAI has no moat" arguments will only make sense whenever there's a material, observable (as in not imaginary), shift on their market share.

[roncesvalles]

>What is OpenAI's first-mover moat?

The same one that underpins the entire existence of a little company called Spotify: I'm just too lazy to cancel my subscription and move to a newer player.

[lotsofpulp]

Not exactly a good sign for OpenAI considering Spotify has no power to increase prices enough such that it can earn a decent profit. Spotify’s potential is capped at whatever Apple/Amazon/Alphabet let them earn.

[aurareturn]

OpenAI has a lot more revenue than Claude.

Late in 2024, OpenAI had $3.7b in revenue. Meanwhile, Claude’s mobile app hit $1 million in revenue around the same time.

[apwell23]

> Late in 2024, OpenAI had $3.7b in revenue

Where do they report these ?

edit i found it here https://www.cnbc.com/2024/09/27/openai-sees-5-billion-loss-t...

"OpenAI sees roughly $5 billion loss this year on $3.7 billion in revenue"

[pradn]

Brand - it's the most powerful first-mover advantage in this space.

ChatGPT is still vastly more popular than other, similar chat bots.

[kpennell]

almost everyone I know is the same. 'Claude seems to be better and can take more data' is what I hear a lot.

[ransom1538]

I moved 100% over to deepseek. No switch cost. Zero.

[qingcharles]

These things aren't the same, though... yet.

ChatGPT is somewhat less censored (certainly on topics painful to the CCP), and GPT is multi-modal, which is a big selling point.

Depends on your use-case, of course.

[ed]

One moat will eventually come in the form of personal knowledge about you - consider talking with a close friend of many years vs a stranger

[kgc]

Couldn't you just copy all your conversations over?

[itissid]

OpenAI does not have a business model that is cashflow positive at this point and/or a product that gives them a significant leg up in the same moat sense Office/Teams might give to Microsoft.

[aurareturn]

Companies in the mobile era took a decade or more to become profitable. For example, Uber and Airbnb.

Why do you expect OpenAI to become profitable after 3 years of chatgpt?

[physicsguy]

Interest rates have an effect too, Uber and Airbnb were starting in a much more fundraising friendly time.

[Der_Einzige]

High interest rates are supposed to force the remaining businesses out there to be profitable, so in theory, the startups of today should be far faster to profitability or they burn out.

[physicsguy]

True, but it makes it much more difficult to get started in the first place.

[meiraleal]

Nobody expects it but what we know for sure is that they have burnt billions of dollars. If other startups can get there spending millions, the fact is that openai won't ever be profitable.

And more important (for us), let the hiring frenzy start again :)

[aurareturn]

They have a ton of revenue and high gross margins. They burn billions because they need to keep training ever better models until the market slows and competition consolidates.

[fairity]

The counter argument is that they won't be able to sustain those gross margins when the market matures because they don't have an effective moat.

In this world, R&D costs and gross margin/revenue are inextricably correlated.

[lxgr]

> First mover advantage acquired and keeps subscribers.

Does it? As a chat-based (Claude Pro, ChatGPT Plus etc.) user, LLMs have zero stickiness to me right now, and the APIs hardly can be called moats either.

[distances]

If it's for mass consumer market then it does matter. Ask any non-technical person around you. High chance is that they know ChatGPT but can't name a single other AI model or service. Gemini, just a distant maybe. Claude, definitely not -- I'm positive I'm hard pressed to find anyone in my technical friends who knows about Claude.

[xmodem]

They probably know CoPilot as the thing Microsoft is trying to shove down their throat...

[jaynate]

They also burnt a hell of a lot more cash. That’s a disadvantage.

[Mistletoe]

I feel like AI tech just reverse scales and reverse flywheels, unlike the tech giant walls and moats now, and I think that is wonderful. OpenAI has really never made sense from a financial standpoint and that is healthier for humans. There’s no network effect because there’s no social aspect to AI chatbots. I can hop on DeepSeek from Google Gemini or OpenAI at ease because I don’t have to have friends there and/or convince them to move. AI is going to be a race to the bottom that keeps prices low to zero. In fact I don’t know how they are going to monetize it at all.

[tw1984]

> DeepSeek just further reinforces the idea that there is a first-move disadvantage in developing AI models.

you are assuming that what DeepSeek achieved can be reasonably easily replicated by other companies. then the question is when all big techs and tons of startups in China and the US are involved, how come none of those companies succeeded?

deepseek is unique.

[11101010001100]

Deepseek is unique, but the US has consistently underestimated Chinese R&D, which is not a winning strategy in iterated games.

[rightbyte]

There seem to be a 100 fold uptick in jingoists in the last 3-4 years which makes my head hurt but I think there is no consistent "underestimation" in academic circles? I think I have read articles about the up and coming Chinese STEM for like 20 years.

[coliveira]

Yes, for people in academia the trend is clear, but it seems that WallStreet didn't believe this was possible. They assume that spending more money is all you need to dominate technology. Wrong! Technology is about human potential. If you have less money but bigger investment in people you'll win the technological race.

[rightbyte]

I think Wall Street is in for surprise as they have been profiting from liquidating the inefficiency of worker trust and loyalty for quite some time now.

It think they think American engineering excellence was due to neoliberal inginuenity visavi the USSR, not the engineers and the transfer of academic legacy from generation to generation.

[coliveira]

This is even more apparent when large tech corporations are, supposedly, in a big competition but at the same time firing thousands of developers and scientists. Are they interested in making progress or just reducing costs?

[corimaith]

What does DeepSeek or really High Flyer do that is particularly exceptional regarding employees? HFT and other elite law or Hedge funds are known to have pretty zany benefits.

[Der_Einzige]

Orwellian Communism is the opposite of investing in people.

[coliveira]

Whatever you think about the Chinese system, they educate hundreds of thousands of engineers and scientists every year. That's a fact.

[11101010001100]

Precisely. This is the view from the ivory tower.

[corimaith]

That doesn't the calculus regarding the actions you would pick externally, in fact it only strengthens the point for increased tech restrictions and more funding.

[rightbyte]

Unique, ye, but isn't their method open? I read something about a group replicating a smaller variant of their main model.

[ghostzilla]

Which brings the question, if LLMs are an asset of such strategic value, why did China allow the DeepSeek to be released?

I see two possibilities here, either that the CCP is not that all-reaching as we think, or that the value of the technology isn't critical, and that the release was further cleared with the CCP and maybe even timed to come right after Trump's announcement of American AI supremacy.

[creato]

I really doubt there was any intention behind it at all. I bet deepseek themselves are surprised at the impact this is having, and probably regret releasing so much information into the open.

[fairity]

It's early innings, and supporting the open source community could be viewed by the CCP as an effective way to undermine the US's lead in AI.

In a way, their strategy could be:

1) Let the US invest $1 trillion in R&D

2) Support the open source community such that their capability to replicate these models only marginally lags the private sector

3) When R&D costs are more manageable, lean in and play catch up

[rightbyte]

It is hard to estimate how much it is "didn't care", "didn't know" or "did it" I think. Rather pointless unless there are public party discussion about it to read.

[lenerdenator]

It will be assumed by the American policy establishment that this represents what the CCP doesn't consider important, meaning that they have even better stuff in store. It will also be assumed that this was timed to take a dump on Trump's announcement, like you said.

And it did a great job. Nvidia stock's sunk, and investors are going to be asking if it's really that smart to give American AI companies their money when the Chinese can do something similar for significantly less money.

[shangofox]

I mean, it's a strategic asset in the sense that it's already devalued a lot of the American tech companies because they're so heavily invested in AI. Just look at NVDA today.

[jerjerjer]

We have one success after ~two years of ChatGPT hype (and therefore subsequent replication attempts). That's as fast as it gets.

[boringg]

Your making some big assumptions projecting into the future. One that deepseek takes market position, two that the information they have released is honest regarding training usage, spend etc.

Theres a lot more still to unpack and I don’t expect this to stay solely in the tech realm. Seems to politically sensitive.

[meiraleal]

DeepSeek is profitable, openai is not. That big expensive moat won't help much when the competition knows how to fly.

[aurareturn]

DeepSeek is not profitable. As far as I know, they don’t have any significant revenue from their models. Meanwhile, OpenAI has $3.7b in revenue last reported and has high gross margins.

[meiraleal]

tell that to the stock market then, it might change the graph direction back to green.

[aurareturn]

I’m doing the best I can.

[WiSaGaN]

Deepseek inference API has positive margin. This however does not take into account R&D like salary and training cost. I believe OpenAI is the same in these aspects, at least before now.