This is a 9amBusiness interfview with David Dobrin, Founder of OriginalVoices
Can you please share more about your current role and background?
I’m the CEO and Co-Founder of OriginalVoices, a platform that builds Digital Twins of real people so businesses and AI tools can understand audiences and test ideas against human opinion on demand. Clients can speak directly to real people through surveys or AI-moderated interviews, query their Digital Twins, or combine the two. Our hybrid model means businesses can do more of both: speak to real people more often and query their Twins more often.
Since launching in December 2025, we’ve grown to more than 50,000+ Digital Twins, announced a data and product partnership with YouGov, and raised £1m in a pre-seed funding round led by Iona Star.
I’ve spent my career at the intersection of consumer data and technology. I previously co-founded Inconvo, a conversational polling platform that was acquired by YouGov in 2018. I subsequently led YouGov’s AI strategy as Product Director of its AI Platform, before leaving last year to build OriginalVoices.
Why have you set up the business?
I started OriginalVoices because I believe that in the age of AI, what real people think and how they see the world should matter more, not less. At the same time, I was seeing the emergence of purely synthetic data solutions attempting to replicate human opinions and perspectives without involving real people at all. I questioned how accurate or representative these approaches could really be, and frankly found the direction quite dystopian.
I wanted to build something different. A way for businesses and AI agents to understand and predict what real people think and feel, grounded in real human data and involving the people themselves.
Our fundamental belief is that AI should amplify and augment real people, not replace them.
What have been the biggest challenges you’ve faced?
One of the biggest challenges has been building trust in an entirely new way of understanding people. AI can generate incredibly convincing answers, but that doesn’t necessarily mean those answers reflect what real people actually think or feel. For us, the challenge is not just building Digital Twins, but demonstrating that they can reliably represent the people behind them.
That’s why we’ve put so much emphasis on grounding our technology in real human data and developing ways to measure and validate its accuracy.
The other challenge is the sheer pace at which AI is evolving. We’re building a business in a market that’s changing almost weekly. That creates enormous opportunities, but it also means constantly adapting, making decisions quickly and staying focused on the things we believe will matter in the long term.
Thoughts on recent suggestions AI could wipe out humans?
AI has enormous potential for both benefit and harm, and I think we should take the risks seriously. But it’s important to distinguish between risks we can observe and measure today and more speculative scenarios about what might happen in the future. My concern is that sensationalised narratives about AI wiping out humanity can sometimes distract from the practical challenges we should already be addressing, such as reliability, accountability, misinformation, and ensuring AI systems remain aligned with human interests.
We need a scientific, evidence-based conversation that neither dismisses the potential risks nor overstates what we actually know.
For me, a fundamental part of getting AI right is making sure humans remain central to how these systems are developed and used. That’s very much the thinking behind OriginalVoices: using AI to amplify human perspectives rather than replace them.
And the rowing back on progress from AI leaders?
I think it’s important to distinguish between progress in AI technology and its adoption across the wider economy. AI capabilities have advanced at an extraordinary pace, but it inevitably takes time for those advances to translate into meaningful productivity gains and economic impact. We’ve seen this with previous technological revolutions. Developing the technology is one thing, but businesses then need to figure out how to integrate it into their operations, redesign workflows and adapt the way people work.
I don’t think that the gap between technological capability and real-world adoption means AI is overhyped or that progress is slowing. It’s a natural part of how transformative technologies spread through the economy. The real opportunity now is to bridge that gap, moving beyond impressive demonstrations of what AI can do at the frontier towards widespread adoption that delivers measurable value for businesses and society.
What commercial difference is AI currently making for companies?
I think one of the biggest changes is the speed at which businesses can go from an idea to a decision.
Take market research as an example. Traditionally, understanding what an audience thinks about a new product, campaign or idea might involve commissioning research, recruiting participants, conducting interviews and analysing the findings over several weeks.
With OriginalVoices, businesses can now explore questions with Digital Twins of real people on demand, test different ideas, refine them and ask follow-up questions in minutes. And importantly, they can still go back to the real people behind those Twins whenever they need to conduct fresh research or validate what they’re seeing.
That combination of real human insight and the speed and scalability of AI is incredibly powerful. It means businesses can make better-informed decisions, more frequently and at a fraction of the traditional time and cost.
How should companies be implementing AI across core functions?
My advice would be to start with the problem you’re trying to solve, rather than the technology itself. There’s a tendency for businesses to feel they need an AI strategy for everything. But the real value comes from identifying specific problems where AI can genuinely improve speed, quality or decision-making.
The other thing I’d emphasise is the importance of data and context. AI models are incredibly capable, but their outputs are only as useful as the information they’re grounded in. And finally, keep humans involved. Whether you’re developing products, creating marketing campaigns or making strategic decisions, AI should help you understand people better, not make assumptions about them.
For me, the biggest opportunities come from combining AI’s ability to process information and operate at scale with the knowledge, experiences and judgement of real people.
What is the vision for the business?
Our ambition is for OriginalVoices to become the human insight infrastructure for AI. As AI agents increasingly take on work on behalf of businesses and individuals, they’re going to need a reliable way to understand what real people think, feel and want. Today’s AI models are incredibly knowledgeable, but they don’t necessarily understand human preferences, experiences and opinions, particularly as those things change over time.
We want to close that gap by building a global network of Digital Twins, each grounded in the data and experiences of a real person, that any business or AI agent can query to understand human perspectives on demand. Our recent £1m pre-seed investment, led by Iona Star, will help us expand that network and continue developing the technology. Ultimately, our vision is simple: to ensure that as AI becomes more powerful and more widely adopted, what real people think and feel matters more, not less.