Piers Linney is not just a former Dragon’s Den investor and entrepreneur — he’s a recognised voice in tech, AI, and innovation. As Co-founder of Implement AI and senior strategist, he blends startup vision with pragmatic insights about emerging technologies.
In this Q&A with The AI Speakers Agency, Piers explores how an AI-first economy will reshape work, how businesses can reduce bias using algorithms, why cybercriminals are already weaponising AI, and how companies should build policies that fit their industries.
Question 1. From your perspective, how will an AI-first economy transform organisations and the way we work?
Piers Linney: “So we, as a fund, a company called Implement AI, we’ve got a framework where urban companies understand what that continuum looks like. And it is a continuum. We call it AI-assisted.
“For 250 years, since the pre-industrial revolution, humans used various tools. But it was basically human-first. Really, we were the ones — it was our cognitive labour, especially. Technology displaced much of our physical labour, but cognitively it was very much human-first.
“Now we’ve entered an age where AI… quite a lot — AI’s been around for 50 years, machine learning in terms of research maybe 10 years, but since the beginning of this year, probably the end of last year, these new large language models, diffusion models, have changed the way we can communicate with technology.
“So what you’re seeing now is a period where we are going to become — and by “we” I mean us personally and organisations — AI-assisted. In that period, software-first, cognitive labour, knowledge
work, the information economy, you’re going to find that your organisations, your employees, can be super-powered.
“Robotics is slightly different. Eventually artificial intelligence is going to go mobile. What we often get confused about is the timeframe. It’s very hard to see more than maybe two years — now it’s like a fog. But what you can be assured of is that there is huge change coming.
“This is not moving from mainframe to network PCs. This is not moving to the cloud. You could take 10 years on that journey if you want to. This is happening very quickly. And we always say that it won’t be your business or your job that’s ruined or out-competed by artificial intelligence — it’ll be your competition using it.
“So what we’re trying to do is say, look, you’ve got two to five years to get on board and embed this. Because what happens is that it’s like a ship leaving the harbour. The gap you can jump now, but as that ship begins to move away at exponential speed — and we’re going to see exponential change over time — that leap is going to be harder to make, to the point where it’s very difficult to make it. And once you are left on the quayside in this new world, you’re staying there.
“Robotics is slightly different. You’re going to see artificial intelligence help humans to design better robotics. They’re quite expensive now, maybe a million pounds for a decent robot, but eventually there’ll be consumer robots that become cheaper through economies of scale. That’s further out, maybe five to fifteen years, but that’s also coming.
“And then eventually we get to a world where it’s AI-first, where everything is driven by AI. You will not trust a human. In fact, I advise quite a large automotive manufacturer, one of the largest in the world, quite recently, and it probably will be illegal for humans to drive a car in a built-up area.
“You won’t trust your human lawyer or accountant. Initially you’ll want them to check something maybe for certain reasons, but it’s going to be AI-first. Humans will have far more faith in artificial intelligence in the future than in our own fallible selves.”
Question 2. Bias has long been a challenge in both hiring and corporate decision-making. How can AI help companies move towards more objective outcomes?
Piers Linney: “AI raises a big issue that people talk about, which is bias. There are different types of bias.
“There’s bias built into these large language models. If you think about ChatGPT or Claude — lots of them, there are open-source ones too — they’ve been trained essentially, simplistically, on the internet, on human data that’s available, including things like Twitter and Reddit.
“So: human bias. The output of humans. And bias is embedded into everything we do. There are over a hundred different types of human bias. So if you’re training something on that content, then you’re probably going to find that some of that bias finds its way into these platforms, into these large language models.
“What they’re trying to do, and working really hard on, is to eradicate bias. There are some platforms, like Perplexity for example, where they’ll give you results but also give you all the references, so you can check where it’s coming from, whether bias is built in.
“But eventually — the beauty of it, I can flip the other way around — is that it enables us to eradicate bias. Because take, for example, when you’re employing somebody. The big issue is unconscious bias. I advise Sky on diversity and inclusion and how to eradicate that.
“Now, with technology, you can. Because you can fine-tune the models to eradicate bias. So when you’re having a conversation with someone you want to hire, potentially, that’s taken out.
“But the big one, related to your question, is business decisions. How do you take human bias out of business decisions? What AI allows us to become — what AI is essentially down the line — is a ruthless optimiser of everything we do, both personally (on personal AIs) and in corporate organisations.
“The corporate AI, eventually, those two are going to talk to each other. And hopefully in those communications — this is why the website and SEO might go away — because AI has to talk to AIs. It won’t have to go and trawl over websites. But eventually even bias out of purchasing will go away.
Question 3. Cybersecurity leaders warn that criminals are already leveraging AI. Why do you believe AI will become such a powerful tool for cybercrime?
Piers Linney: “Artificial intelligence can be used for good, but it can also be used by bad actors, and increasingly it already is. One of the battlefields we’re going to see is cyber.
“I did a talk quite recently for one of the UK’s, in fact the world’s largest — a top-five cyber company. I was talking to senior management there and we were discussing the cyber arms race.
“I said that a lot of this malware in the future is going to be designed by AI. It’s going to be very sophisticated. And they looked at each other and said, “It’s actually worse than that, Piers. The malware will be an AI.”
“So you’ll have some intelligence, essentially something which has an objective and can sit inside your systems, your computers, your watch even. It can sit there with its objective and then work out a way to achieve that objective over time, even before you may be aware of it.
“So we’re going to see this battle. This was always going to happen to all technology — it’s nothing new. But when you start talking about artificial intelligence, and maybe quantum computing, then you’re getting into something quite serious.
“The main danger, I think, out of AI is not the future of robotic-looking humanoid androids crunching over the skulls of defeated human armies. It’s more about the fact that cybercriminals can use this in very dangerous and nefarious ways to harm people, especially financially — not just people, but potentially whole countries and governments.”
Question 4. Every industry is wrestling with AI adoption differently. What practical steps should companies take when designing an AI policy tailored to their sector?
Piers Linney : “AI policy is something that every company needs to embed. At my business, Implement AI, this is something we actually do.
“It’s kind of an overall framework you can embed into your business. And it depends what industry you’re in. It can be quite different in regulated industries — for example, in healthcare, in terms of data. So you have to have an overall framework.
“The key is that it pervades everything. You can’t just have a standalone one. What you’ll find is that aspects of that will need to be embedded into all of your policies — everything from HR to maybe even holiday policy.
“So, a standalone one probably isn’t the way to go. It’s something which pervades the whole business. Because essentially, artificial intelligence will — and you may not know it — but employees now, research says that 30% of employees are using AI in their day jobs.
“So you’ve got to be very careful that your company data, your personal data, your customer data is being treated correctly, and that people are using this technology in ways that actually add value to your business.
“Also, training is really important. People need to know how to use it. It’s one thing having access to artificial intelligence. It’s another thing actually knowing how to use it.
“People need to know how to use it in a responsible way to impact your business or organisation. If it’s data security, your customer data security, then you can add a huge amount of value to your business.
“Not allowing people to use AI is not the solution — because they’re always going to. It’s called “shadow AI”, like shadow IT. Not having the policy frameworks of what are the four corners of the box in which they can use it — that is a dangerous way forward.
“You need to make sure people know your organisation’s approach and policies for using this extremely powerful technology.” This exclusive interview with Piers Linney was conducted by Mark Matthews of The Motivational Speakers Agency.
Leave a Comment