Sensay’s Dan Thomson: Putting Humanity Back Into AI - BlockTelegraph

* Facebook

• Twitter
• Buffer
• reddit
• LinkedIn Artificial intelligence is often accused of stripping customer conversations down to blunt decision trees. Dan Thomson, co-founder and chief executive of Sensay, insists the opposite is possible.

“We are putting the humanity back into the AI,” he tells me, neatly summing up the task his remote based enterprise has set itself: building always-on digital replicas that sell, serve and moderate as naturally as a seasoned employee .

From philosophy to product-market fit

Thomson’s résumé is unusual even by Web3 standards. A degree in philosophy, years in finance, and a spell running bars and restaurants all feed into his obsession with ‘the theory of mind’ and ‘digital immortality’. When language models matured, he saw his chance to commercialise the ideas first outlined in his early books.

The result is Sensay’s ‘AI brain’, a toolkit that trains a replica on its creator’s voice, style and data, then deploys it as a sales agent, customer-service lead or community manager. Crucially, the system plugs into multiple models, ChatGPT, Claude, Grok and others, while adding a retrieval layer for up-to-date company knowledge.

“ChatGPT is clever, but it does not know you prefer taco trucks to fine dining,” Thomson says.

The business case: revenue up, costs down

Clients are buying in. Thomson cites examples where adding a persona to the website lifted online conversion ‘three-hundred per cent’ and cut support costs by ‘fifty to seventy per cent’.

In industries from higher education to manufacturing, replicating the CEO or head of sales gives prospects a warm, data-rich conversation instead of the familiar chatbot cul-de-sac that ends with the plea, ‘Can I speak to a human being please?’

Community management is an equally strong use case. Telegram groups and Discord servers run all night; humans need sleep. A replica answers in any language, holds the brand tone, and never forgets the latest policy update.

Tokenomics that mirror revenue

Sensay is a Web3 firm, so it carries a token. Last year’s Token Generation Event (TGE) raised more than three million dollars and seeded the utility token, $SNSY. Thomson knows most corporate buyers could not care less about tokenomics, yet he argues the structure matters.

“Buying pressure comes directly from our revenues,” he says. A slice of every licence fee is used to purchase $SNSY on the open market, while rewards are paid to users who build high-quality replicas or refer new customers. Staking yields currently reach 50 percent for simple locks and 120 percent for liquidity mining, numbers that give long-term believers an incentive to hold.

Ethics and regulation

Talk of digital immortality quickly drifts into ethical terrain. Does a faithful replica need to capture ‘the best and the worst’ of a personality? Should it swear? Could it turn toxic if trained on malice and misinformation?

Thomson leans on existing frameworks, GDPR, SOC 2 and emerging AI laws, but concedes they are only a start.

“When you replicate humans you face every ethical boundary at once,” he says. “We think of the replica as an extension of the person, never as mandatory, never as a job killer.”

Data-ownership matters too. Sensay allows users to decide which emails, messages or social feeds feed the engine. Updates are human-in-the-loop; a marketing team can tune tone and facts, then push fresh weights to the model.

The road to digital legacy

Whole-brain emulation is still science fiction, Thomson admits. What Sensay offers today is pragmatic continuity: a replica that can answer staff questions, moderate community channels and, if the user chooses, keep speaking after the flesh-and-blood original is gone.

“It is a tool, an option, something you can pass to your kids if you want,” he says .

For now, the company’s mission is simpler: prove that AI personalities can drive revenue while sounding like real people. In a market flooded with faceless bots, the promise of a sales rep who remembers your last conversation, and cracks a joke in your own slang, may be all the persuasion enterprise buyers need.