>If they IPO'd today, they'd easily be worth hundreds of billions if not more.

You can't just say things like this without justification. The last raise was in October, at $157B. Did OpenAI, the hottest startup in history, give away equity at a discount?

Now, since that time, has the value gone up or down? We now know they have serious technological competition and are losing lots of money.

I think the product is great, but they aren't worth limitless amounts of money.

OpenAI is currently in talks to raise at a $340B valuation.

https://techcrunch.com/2025/01/30/openai-said-to-be-in-talks...

Who in their right mind would put money into OpenAI with DeepSeek out there dumping its better models for free with no license restrictions?

Anyone with enough hardware can run the model and run a competing service for less.

Many companies currently using OpenAI are considering alternative cheaper and better options.

The OpenAI moat has been filled in.

Absolutely, we can even make a parallel with UNIX:

UNIX was revolutionary. Companies poured billions into proprietary UNIX versions: IBM's AIX, Sun's Solaris, HP-UX, SGI IRIX...

Then in 1991, Linus Torvalds released Linux for free.

What was the point of UNIX anymore then ?

Unless OpenAI locks in customers with exclusive data, it risks becoming just another proprietary relic beaten by open alternatives.

You've got a point about UNIX and Linux, but let's not forget about Microsoft. Back in the day, even with free systems popping up, Microsoft became a tech giant with its pricey products. Similarly, OpenAI has the potential to be the Microsoft of AI. With its resources and smart partnerships, it could set the trend and stay ahead, even as open-source options grow.

This would have been true 2 years ago.

Now with DeepSeek-R1, any company can host very similar service, so the technological edge has been lost.

Google itself offers similar experience with Google Flash 2.0 Thinking.

We haven't seen revolution with o3 (most of us, programmers, here are still with claude-sonnet-3.5 than the supposedly groundbreaking gpt-o3).

This. Most users don't care about what model has the best benchmarks on twitter.

They care about the heavy censorship of the model, the amount of refusals and the low quality of the answers with the o1/o3 series.

They seem to be doing ok...

https://trends.google.com/trends/explore?date=now%201-d&geo=...

IPOs are pump and dumps; this much is a given.

OpenAI isn't even profitable yet, and they might never become profitable.