Original Article Title: Owning vs. Renting Intelligence
Original Author: Lin Qiao
Translation: Peggy, BlockBeats

Editor's Note: Mythos was shut down this week, prompting many AI entrepreneurs to reconsider an issue that cost discussions have obscured: when a product's core capability is built on external models and platforms, what does the company truly own?

Over the past few years, open-source models have often been discussed within the framework of "cheaper cutting-edge model substitutes." However, this article argues that cost is not the most crucial variable; control is. For an AI company, using cutting-edge model APIs can quickly launch a product, lower the technical threshold, but also means that the core capability may be subject to the model supplier's rules, pricing, policy adjustments, or even decisions to remove the model.

The article further suggests that "owning intelligence" does not mean abandoning cutting-edge models, but that companies should embed their own data, workflows, domain knowledge, evaluation standards, and edge cases into a controllable model system. The future competition of AI may not necessarily be dominated by a single largest model but will see multiple "frontiers": universal cutting-edge models, enterprise-specific post-training models, vertically specialized models, and routing systems composed of multiple models working in concert.

As a result, Mythos's shutdown serves as a reminder: in the AI era, the real moat is not just the ability to call upon powerful models but whether intelligence can be transformed into a company's own asset.

The following is the original text:

Mythos was shut down this week. Whether you agree with this decision is no longer the focal point.

What truly struck many people is that a company built on intelligence it cannot control suddenly found itself exposed to decisions it couldn't influence. Many founders, upon witnessing this scene, ask themselves the same question: which parts of my business are actually just "rented"?

Over the past few years, discussions about open-source models have mostly revolved around cost: can they really get the job done? If so, how much cheaper are they compared to calling upon cutting-edge model APIs?

Now, we have a fairly clear answer. We have worked with companies like @RampLabs, @cursor_ai, @harvey, following a similar playbook: starting from a powerful open-source model, conducting post-training on work content that truly matters to the company, and continuously benchmarking it rigorously against cutting-edge models.

The results have been surprising time and time again. When it comes to the most critical tasks for businesses, a fine-tuned open-source model can often achieve or come close to the quality of cutting-edge models at a very low cost.

But what has truly become clear this week is this: cost has never been the most significant issue.

The deeper issue at hand is control. Who owns the intelligence that your product relies on?

Many recent discussions have been framed as the difference between "renting" and "owning." This analogy is not perfect but is quite useful.

Renting Intelligence

Renting has been working well until things go wrong. Apartments are move-in ready with functioning lights, water, and maintenance services. That's why most companies initially choose this path.

Cutting-edge model APIs are excellent products. They enable startups to build things that seemed unimaginable just a few years ago.

However, renting also means limitations. Landlords can raise rent, decide what improvements you can make, change the rules, and occasionally, for reasons unrelated to you, they can tell you it's time to leave.

You haven't done anything wrong. You've simply been operating on someone else's turf.

That's why Mythos's story has resonated with so many people. When your core capability relies entirely on someone else's platform, you are exposed to a set of decisions that are out of your control.

Most of the time, this isn't critical. But sometimes, it can become extremely important in an instant.

Owning Intelligence

The lesson here is not that companies should stop using cutting-edge models. Far from it. The labs producing cutting-edge models have created remarkable technology. Most products should use them. We do too.

In many ways, cutting-edge models are becoming infrastructure. However, infrastructure and ownership are two different things.

You can use public infrastructure while still owning something that truly creates value for your business. In the field of AI, "owning" means starting from a state-of-the-art open-source model and shaping it around the most unique parts of your company.

Your data.

Your workflows.

Your domain knowledge.

Your edge cases.

Your Benchmark.

Your Definition of "Good."

Over time, this model will become less general and more reflective of the actual work your company does every day. Value is created right here.

Think of it like a house. It's easy to move the furniture around, easy to paint a wall. But if your future depends on the layout of the house itself, you'll eventually want the ability to move walls. Intelligence works the same way.

When intelligence truly belongs to you, no one can quietly take the floor out from under your product.

That's why we're building Fireworks this way.

We've combined training and inference into one system, allowing companies to adopt the best open-source models, shape them around their most critical business problems, and deploy them stably into production.

Not just consuming intelligence. But owning intelligence.

No Single Frontier

There was also an optimistic revelation this week: the future of AI does not depend on one model winning it all.

There is no single frontier. There are many.

A frontier model is one frontier.

A model retrained on years of proprietary company knowledge is another frontier.

A specialized model that outperforms any model on a narrow problem is yet another frontier.

A system that can route requests to multiple models, get them to cooperate, and surpass a single model across many tasks is also a frontier.

The most interesting shift in AI isn't that one model is getting smarter and smarter, but that intelligence is becoming more and more customizable.

The winning companies in the end won't necessarily be the ones with the biggest models, but those who can turn intelligence into their own unique asset.

Looking Ahead

Much of this week has been spent reacting to the news, and we've chosen to continue releasing products: @Kimi_Moonshot K2.7 Code, @MiniMax_AI M3, @Alibaba_Qwen 3.7 Plus.

What I look forward to in the future is not a model quietly consuming all it sees.

Instead, many teams can have their own piece of the cutting edge.

If Mythos being shut down has you rethinking trade-offs, we'd love to chat.

[Original Tweet]

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia