Short answer: no. Not all of them.

We are in a phase where AI capability is improving quickly, but business fundamentals still apply. If a company loses money on every heavy query, pays escalating compute bills, and has no durable distribution, the runway eventually ends. Capital markets can delay that outcome, but they cannot repeal arithmetic.

Why losses feel normal right now

Many AI companies are deliberately in investment mode. That part is not surprising. Early infrastructure waves often look unprofitable before pricing, hardware, and product shape settle.

The issue is not "losses exist." The issue is unit economics that do not visibly improve.

The survival filter is becoming obvious

Over the next few years, the winners are likely to have at least two of these three:

• Distribution: default placement in existing workflows (office suites, code platforms, clouds, devices).
• Cost control: a credible path to cheaper inference per useful task.
• Monetisation power: users or enterprises willing to pay enough to cover real usage.

Companies with one flashy model and no channel will struggle. Model quality alone is becoming less of a moat as capabilities converge and open weights spread.

The contrarian point

Most people say model intelligence will decide everything. I think that is backwards.

The biggest moat in AI will be distribution plus trust, not benchmark leadership. The company embedded in day-to-day work with acceptable quality and reliable governance will often beat the technically superior outsider.

What to watch instead of hype cycles

• Gross margin trends on AI products (not just revenue growth)
• Attach rates inside existing products
• Enterprise renewal quality
• How often users return without incentives
• Whether pricing survives contact with heavy usage

There will be big winners in this cycle. But there will also be quiet failures: teams that built impressive demos, raised aggressively, and never escaped expensive inference with weak distribution.

AI is not exempt from business gravity. The companies that survive losses are the ones that can explain when and how those losses convert into durable cash flow.