When the Team Is the Traction: Is AI Changing What Investors Back?

Some companies are attracting extraordinary amounts of capital before they have the conventional evidence we associate with an investable startup.

One of the most striking examples emerged this month.

British AI startup Emulate, founded in August by three former Google DeepMind researchers, is reportedly in discussions to raise as much as $700 million at a $3.7 billion valuation.

At the time the funding talks were reported, the company was barely a month old and reportedly had no product, revenue or public website. Financial Times

So what exactly are investors backing?

Increasingly, at the very edge of AI, the answer may be the people themselves.

When Experience Becomes an Asset

Emulate's founders aren't typical first-time entrepreneurs.

Jack Parker-Holder, Matthew McGill and Philip Ball previously worked on Google DeepMind's Genie programme, developing so-called "world models": AI systems designed to understand and simulate environments rather than simply generate text.

That experience is exceptionally scarce.

And investors appear prepared to place an enormous value on it. 

In a conventional startup, traction helps reduce uncertainty.

Revenue proves somebody will pay.

Customer growth demonstrates demand.

Retention suggests the product has value.

But what happens when a company is trying to build technology so new that conventional commercial traction doesn't yet exist?

Investors may begin looking elsewhere for evidence.

Who are the founders?

What have they built before?

What technical problems have they already solved?

And perhaps most importantly: are these among the relatively small number of people in the world capable of building what comes next?

The Billion-Dollar Seed Round

Emulate isn't an isolated example.

Earlier this year, London-based Ineffable Intelligence, founded by former DeepMind researcher David Silver, raised $1.1 billion in seed funding at a $5.1 billion post-money valuation.

The British Business Bank described it as the largest European seed round in history and invested $20 million alongside investors including Sequoia Capital, Lightspeed, NVIDIA, Google and others. British Business Bank

Silver's background helps explain why.

He previously led reinforcement-learning research at DeepMind and contributed to breakthroughs including AlphaGo and AlphaZero. Ineffable is now attempting to develop AI systems capable of learning through experience rather than relying primarily on existing human-generated data. British Business Bank

Again, investors weren't backing years of company revenues.

They were making an enormous bet on technical capability and what a particular team might be capable of creating.

The Rise of the AI 'Neo-Lab'

A new type of company is emerging around this phenomenon.

Sometimes described as neo-labs, these businesses are being created by researchers leaving organisations such as DeepMind and other leading AI laboratories to pursue new approaches independently.

The economics can look very different from those of a traditional software startup.

Training frontier models can require vast amounts of computing power. Elite researchers command exceptional salaries. Infrastructure costs can run into hundreds of millions.

That means the capital required simply to discover whether an idea works can be enormous.

And because competition for the strongest AI researchers is intense, investors may feel they cannot wait for conventional traction before investing.

By the time the product exists, everyone else may already want in.

Does That Mean Traction Doesn't Matter Anymore?

Absolutely not.

And this distinction is important.

The extraordinary funding rounds being raised by frontier AI laboratories should not become the new benchmark by which ordinary startups judge themselves.

A founder building an AI-enabled accounting platform, recruitment product or healthcare service will still normally need to demonstrate a compelling commercial opportunity.

Investors will still ask about customers.

Revenue.

Competition.

Margins.

Distribution.

Retention.

And whether AI genuinely gives the business an advantage rather than simply appearing in the pitch deck.

The companies raising hundreds of millions before launching products are exceptional precisely because the founders themselves have unusually strong evidence behind them.

Their previous work is part of the traction.

For most companies, it isn't a substitute for it.

Not Every AI Company Is an AI Lab

There's another distinction investors increasingly need to make.

Calling a company an "AI startup" tells us surprisingly little.

At one extreme are frontier laboratories attempting to create entirely new AI architectures and models.

At the other are thousands of businesses using existing AI technology to solve specific commercial problems.

Both can create valuable companies.

But they require very different investment cases.

A frontier laboratory may require enormous amounts of capital, scarce scientific talent and years of technical development.

An application-layer startup might instead succeed because it understands a particular customer better, owns valuable proprietary data or has found an effective way to distribute an AI-enabled product.

For investors, understanding where the company's actual advantage sits is becoming increasingly important.

Simply using AI isn't a moat.

A Bigger Opportunity for British AI

There is also a wider UK story here.

The Government is now putting significant capital behind keeping ambitious AI companies in Britain. Its Sovereign AI programme represents a £500 million commitment to backing homegrown AI companies, including early-stage equity investment, access to computing infrastructure and other support. GOV.UK

In August, the Government also announced a £100 million procurement programme designed to give British AI startups opportunities to develop technology addressing challenges in areas including healthcare and national security. GOV.UK

Alongside private investment, that creates an increasingly substantial ecosystem around British AI.

And the emergence of companies such as Ineffable Intelligence and Emulate suggests something particularly valuable is already here:

talent.

Britain isn't simply trying to attract AI companies from elsewhere. Some of the researchers who helped create major breakthroughs in modern AI are now founding their next companies here.

What Investors Are Really Buying

Every early-stage investment contains an element of belief.

There is never enough data to know exactly what a young company will become.

Investors therefore look for signals.

A strong market.

Early customers.

Revenue.

Intellectual property.

Technology.

Founder experience.

What the current generation of frontier AI companies demonstrates is that the weighting of those signals can change dramatically depending on what is being built.

For some of the most technically ambitious AI businesses, the people building the technology may be the strongest signal available.

That doesn't mean the old rules of startup investing have disappeared.

It means that at the frontier of a rapidly developing technology, the definition of traction can occasionally look very different.

And when only a small number of people in the world have already demonstrated that they can build breakthrough AI systems, investors may decide that waiting for conventional proof is itself the bigger risk.