undefined | Better HN

0 pointsinfecto5mo ago0 comments

Does DeepSeek have any market penetration in the US? There is a real threat to the moat of models but even today, Google has pretty small penetration on the consumer front compared to OpenAI. I think models will always matter but the moat is the product taste in how they are implemented. Imo from a consumer perspective, OAI has been doing well in this space.

0 comments

rdtsc5mo ago

> Does DeepSeek have any market penetration in the US?

Does Google? What about Meta? Claude is popular with developers, too.

Amazon? There I am not sure what they are doing with the LLMs. ("Alexa, are you there?"). I guess they are just happy selling shovels, that's good enough too.

The point is not that everyone is throwing away their ChatGPT subscriptions and getting DeepSeek, the point is that DeepSeek was the first indication the moat was not as big as everyone thought

infectoOP5mo ago

Maybe my point went over the fence.

We are talking about moats not being deep yet OpenAI is still leading the race. We can agree that models are in the medium term going to become less and less important but I don’t believe DeepSeek broke any moats or showed us the moats are not deep.

brandon2725mo ago

> The point is not that everyone is throwing away their ChatGPT subscriptions and getting DeepSeek

Currently.

infectoOP5mo ago

Certainly but also the OP was making Ana argument that DeepSeek proved something and I am arguing that most of this market is not valuing on the model but the consumer commercial offering.

rdtsc5mo ago

Exactly, right? It went from "OpenAI seems like years ahead" to "well, maybe if DeepSeek can do it so can we, let's try".

1 more reply

j / k navigate · click thread line to collapse

0 comments

rdtsc5mo ago

> Does DeepSeek have any market penetration in the US?

Does Google? What about Meta? Claude is popular with developers, too.

Amazon? There I am not sure what they are doing with the LLMs. ("Alexa, are you there?"). I guess they are just happy selling shovels, that's good enough too.

The point is not that everyone is throwing away their ChatGPT subscriptions and getting DeepSeek, the point is that DeepSeek was the first indication the moat was not as big as everyone thought

infectoOP5mo ago

Maybe my point went over the fence.

brandon2725mo ago

> The point is not that everyone is throwing away their ChatGPT subscriptions and getting DeepSeek

Currently.

infectoOP5mo ago

Certainly but also the OP was making Ana argument that DeepSeek proved something and I am arguing that most of this market is not valuing on the model but the consumer commercial offering.

rdtsc5mo ago

Exactly, right? It went from "OpenAI seems like years ahead" to "well, maybe if DeepSeek can do it so can we, let's try".

1 more reply

j / k navigate · click thread line to collapse