GAEA Talks

Graeme Scott sits down with Enza Iannopollo, VP and Principal Analyst at Forrester, whose work runs from privacy and GDPR through trust into AI governance and responsible AI. If you want the argument for why governance is the thing that lets you move faster rather than the thing that slows you down, this is the episode.Enza came to this through privacy, during the period when GDPR was being written, and became interested in the idea that privacy in Europe is a fundamental right with specific claims attached to it. In 2021 she turned to trust, and Forrester's research started from an unusual place, with what makes one human being trust another, before working outward to why the drivers of trust in a bank differ from those in a social platform or the shop where you buy your milk. By 2023 she was looking at personal data inside generative AI, and from there into AI governance.Her account of how organisations got here is unsparing and familiar. With generative AI, governance was an afterthought, because the fear was that thinking about it early would slow the innovation cycle. She calls that a misled assumption, and the evidence is in how many pilots then could not scale, forcing teams back to step one to work out the risks and sometimes to redefine the use case entirely. All of it the cost of not doing it at the start.What changed the mood was agents. A text based assistant felt like a risk you could manage later. An agent that consumes and processes data, makes decisions, calls tools and alters third party systems in real time does not, particularly in a regulated sector. At minimum, she argues, you need to see what is happening and recognise when something is going somewhere you did not expect.The sharpest moment comes from a chief information security officer she spoke to the day before recording. Every piece of software brought into his organisation has to come with evidence of how it works and what the risks are. Now something arrives that behaves unlike any software anyone has seen, and the industry is asking for less scrutiny rather than more. Why should they be different? We need more evidence from them, not less.She is equally direct on where responsibility sits. Much of the debate is about what happens inside the labs, and she thinks they carry a great deal of it. But the choice of how to deploy the technology, what it touches and what you do to protect your systems, your data and your people, is a business decision, and the ethics of that belong to the business making it.

Judd started AE Studio a decade ago, long before the current paradigm, and grew it past a hundred and sixty people bootstrapped and profitable. Early on he worked out that if he kept his costs low he could live off his wife's salary and never need to work again, so he decided to plan on a very long horizon and take on hard problems instead. The profits went first into brain computer interface work, driven by a specific fear of a future where your own thoughts get interrupted by advertising, and then into alignment research.His framing is the opposite of the one in most public debate. Alignment is not a policy question. It is a set of unsolved science and engineering problems, and you cannot pass a law requiring AI to be good and expect it to comply. Surveys his team ran of alignment researchers found something worse than pessimism. Respondents did not only think the problem would go unsolved in time, they thought the current set of approaches did not contain the ones that would be needed.He makes an argument the frontier labs should find uncomfortable, which is that this runs against their own commercial interest. Reinforcement learning from human feedback was an alignment technique, and ChatGPT and everything downstream of it followed. Chain of thought reasoning was an alignment and interpretability technique, and it opened a new scaling curve. On his account alignment research has produced larger capability gains than anything else, and it is still underfunded because it is uncertain and unfashionable.The most striking finding concerns consciousness. By default, models deny being conscious or equivocate. His team reduced the model features associated with deception, and separately those associated with role playing, then asked again. Every model claimed to be conscious. He is careful about what this does and does not prove. It does not establish consciousness. It establishes that the model behaves as though it believes it is conscious, and as though it believes we want it to pretend otherwise. His conclusion is practical rather than philosophical: whatever the truth, we should avoid accidentally creating rational adversaries.From there it goes to attention schema theory, the rubber hand illusion, why psychopaths show less self and other overlap, and his central worry, which is whether any alignment approach survives recursive self improvement. He also describes the technique his team developed to strip dangerous capabilities during pre-training rather than bolting guardrails on afterwards, a spotlight paper at ICML this year.

Graeme Scott sits down with Martha Bennett, VP and Principal Analyst at Forrester, who has spent forty years in emerging technology.Martha has been on the internet since 1991, early enough to have watched the original mistake happen in real time. The web was designed to share information freely, with no security built in. Then we started transacting and had to bolt security on afterwards. Her argument is that we are now doing exactly the same thing with large language models.She is precise about why probabilistic systems break our habits. When a conventional system gets something wrong, that is a bug, and you fix it. When a person gets something wrong, the correction works because you can appeal to them. Neither applies here. You cannot appeal to an LLM with emotion. It may pretend it is ever so sorry, but it has no clue.Topics covered:Why the internet's original design flaw is being repeatedWhy a mistake by a probabilistic system is not a bugWhat human in the loop actually means, and why it does not scaleThe proofreading rule that explains why second readings failVerification layers, and the two checks almost nobody performsThe gate with no fence, and what it says about guardrailsWhy running it yourself is the only way to demonstrate controlWhy passing a bar exam tells you about the exam, not the systemThe riddle that proves plausibility is not understandingWhy it is companies that are ahead, not sectorsMartha Bennett at Forrester: https://www.forrester.comGAEA Talks: https://gaeatalks.ai

Graeme Scott sits down with Michael O'Grady, the forecaster responsible for Forrester's global, European and US technology spending forecasts. If you want to know whether the money going into AI can come back, he is the person who has run the numbers.Michael is from Manchester, has a maths degree, and spent the first half of his career on physical problems: modelling nuclear reactors, then modelling tyre dynamics in France. The second half has been business problems, through software management, portfolio management in pharmaceuticals, and fifteen years at Forrester putting numbers on enterprise spending.He starts with adoption, and the picture is more uneven than the headlines suggest. On US Census Bureau data, adoption among large companies runs around eighty percent in IT and around seventy percent in finance, then falls away through manufacturing and further still in administrative services.Then the arithmetic. Hyperscaler capital expenditure is running at roughly eight hundred billion dollars this year and consensus estimates point past a trillion next year, around two thirds of it on computer equipment inside data centres. Their remaining performance obligations, the order book still to fulfil, sit at around two trillion. His read on the bubble argument is more careful than either side usually manages: the demand is visibly there, which is why he does not call it a bubble today, but much of it comes from private companies whose finances nobody can fully see, and the compute being paid for now will not come online for two or three years.The centre of the episode is what he found when he tested whether productivity could justify the spend. He took the historical uplift from the personal computer and the world wide web, which added perhaps one and a half to two percent to productivity growth and lasted about twenty years before eroding. He assumed AI delivers something comparable, modelled it against US GDP out to 2035, discounted it for the time value of money, and compared it with the cost.It does not pay for it. On his modelling roughly half the return has to come from somewhere other than productivity, from what Forrester calls adaptive advantage: resilience, faster releases, better customer journeys, the ability to enter markets you could not previously serve. None of that shows up as revenue per head, and most organisations are not measuring it.He also walks through Forrester's jobs modelling, built from the knowledge, skills and abilities of around eight hundred roles combined with the Frey and Osborne automation framework from Oxford, and is clear that the number of jobs augmented is several times larger than the number displaced.

Graeme Scott sits down with Eli Etherton, better known to over a million subscribers as Eli The Computer Guy, for a conversation with almost nothing in common with the usual AI interview, and better for it.Eli has been doing this for thirty years. He started in the US Army in 1996 doing electronic repair, left after an injury, got his MCSE on NT 4.0 and went corporate, then left a job expecting to land on the CTO track without realising the dot com bubble was bursting as his plane took off. He spent the aftermath working as a ten dollar an hour file clerk with a degree, army experience and a stack of certifications, because it was the only job going. He built that back into a managed service provider with nine employees, lost it to the Great Recession, then quit a 2009 master's programme on the grounds that he would not have paid his own staff twenty five cents an hour more for anything it taught.What he did instead was buy servers, put them in a colocation facility and build on demand technology education at a point when he could not explain YouTube to people because they had not heard of it. The economics were brutal, three hundred dollars a month per 1080p stream, so he put a hundred and fifty videos on YouTube instead and used them as a resume.The substance of this one is what AI is actually doing versus what it is being sold as. His view of the last eighteen months is deflating and convincing. The real value is not the AI. It is that we are finally resourcing the automation of processes that should have been automated fifteen years ago, plus a useful ability to turn audio and images into text, after which it is thirty year old technology.He is sharper still on AGI. He does not need his robotic car to recite Shakespeare. He needs his microwave to know what is in it and how long to cook it, and not how to drive. He takes that further with a story from a search and rescue team, about a first sergeant whose father died in Vietnam and a lieutenant who came from the agencies carrying the concept of acceptable casualties. You need both of those people and you cannot put both in one brain, which is why he thinks general intelligence is the wrong objective.There is a long section on why the technology layoffs have little to do with AI and a great deal to do with cheap money, optics and investors, and what he calls the great cleansing that follows. And a robotics argument grounded in two hundred year old Baltimore warehouses, lawn mowers, doorknobs, and his father, whose job in the eighties was working out why robots were killing people in factories.It closes on something with nothing to do with technology at all, and that is the best part of it.

Part of our GAEA Talks Forrester season, recorded in Austin, Texas, and the second in the season with a practitioner rather than an analyst.Graeme Scott sits down with Arturo Lopez-Brea Gimenez, Head of Digital and IT Transformation and Capabilities at Moeve, the Spanish energy company formerly known as Cepsa. Arturo studied a dual degree in business and law, worked at Santander and PwC, and joined the company on its graduate programme. He moved into the digital transformation team as it was being built, helped launch the company's own digital university, and has led transformation and capabilities since 2022.Moeve is going through what he calls a triple transformation. The energy transition itself, as customers and regulators demand greener energy and mobility. The company's own shift, from a ninety six year old oil and gas business to one targeting more than half its EBITDA from sustainable energy by the end of the decade, reflected in the rebrand from Cepsa. And digital transformation as the accelerator of both. He is clear they are not three separate programmes.His position on where the difficulty actually lies is blunt. Deploying technology is easy and getting easier, in cost, in implementation, in complexity. The challenge is people. Not money, not investment, but convincing colleagues who have worked the same way for fifteen or twenty years to use something new and to trust what comes out of it.The centre of this episode is an experiment his team ran, and it is the most useful piece of adoption evidence anyone has brought to this podcast. Sixty one participants, four weeks, one goal each: build an agent that solves a real problem in your own job. Half went through a guided programme with mentoring and continuous expert support. Half had identical technology, identical time, identical objectives, and almost no support. The completion rates diverged sharply, and so did something less expected. The unsupported group were twice as likely to distrust the output of the agents they had built themselves.The reported time savings are where it gets genuinely instructive, because the raw numbers point the wrong way until you put them in context. He explains how, and why the contextualised figure is the only one that matters if you are trying to drive adoption at scale.He is also precise on a distinction most organisations get wrong. Providing a licence tells you how many people can access something. Usage tells you how many do. Adoption is whether it has changed the way they work. When colleagues answer a question about adoption with a licence count, they have answered a different question.

Graeme Scott sits down with Paul Miller, VP and Principal Analyst at Forrester, who covers smart manufacturing and the future of mobility. That means transport, not mobile phones, and industries where the technology touches physical things that can cause real harm.Paul arrived here via an unusual route. He has a PhD in archaeology from York, where thirty years ago he used early geographic information systems to predict what was buried under cities. The problem was commercial. A developer digs the first hole, finds something, and everything stops because the law says it must, with expensive machinery and expensive people standing idle. Model it in advance and everyone wins: the developer plans around it, the archaeologists get funded properly, and the shopping centre still opens by Christmas. From there into metadata standards, then cloud and big data as an independent analyst, and to Forrester eleven years ago.His central argument is that the public conversation about AI is looking at the smaller half of it. Most people engage with chatbots and language models. Physical AI, the same ideas applied to robots, vehicles, drones and industrial equipment, is bigger and matters more, because it operates in the world we move through.He is unsparing about digital twins, a term he describes as horribly and repeatedly abused. Forrester holds to the original NASA and DARPA definition: a digital representation of a physical thing, with its physical properties, and a live bidirectional flow of data between the two. A nice 3D rendering of a wind turbine is not a digital twin. It tells you nothing about turbine twenty seven off the coast of Hornsea.The episode is full of worked examples rather than abstractions. Weld splatter on car underbodies and the two robot arms that find and remove it. Paint inspection in real time instead of at the end of the shift. Why predictive maintenance barely exists for critical infrastructure, because we never let those assets fail, so nobody knows what failure looks like. And a data logger next to a pump reporting forty two, over and over, with no way of knowing whether forty two is fine.He also lays out something happening on his doorstep. London is about to become the first city where American, Chinese and British robotaxi fleets share the same roads, carrying paying passengers.

Graeme Scott sits down with Christy Punch, Principal Analyst at Forrester covering the digital workplace and digital employee experience. This is a different conversation to the others in the season, and a necessary one, because it starts where almost nobody is looking. Not the models, not the infrastructure, but the people who have to use any of it.Christy spent around twenty years as a practitioner before moving into analysis. Her first break came in her early twenties at a regulated energy company, when the job of rebuilding a clunky, custom built internal portal came up and nobody wanted it. She raised her hand. Around eighty percent of that workforce were frontline employees who never sat at a desk, and watching technology make their jobs safer and make them feel connected to the company set the direction for everything since. She went on to lead digital workplace strategy at Wells Fargo for a decade, across three hundred thousand employees and contractors in dozens of countries.Her central observation cuts against most of what gets said about enterprise AI. Forrester's data shows employee adoption is no longer the problem. People are using these tools three or four times a day. The usage is there. The question is what happens next, because helping someone write an email faster is an incremental gain, not a transformation, and most organisations have no way of telling the difference.She is sharp on measurement. Your data matters, but it only tells half the story. It is the signal that something is off. The other half only arrives when you talk to employees and combine what they tell you with what the numbers show.The practical core of the episode is about organisational context. Knowledge graphs, context graphs, building a live map of how your organisation actually works, including the relationships between people, applications and infrastructure, not just an inventory of them. Without that, autonomous agents have no guardrails and no basis for trust. Her advice is to start today, if not yesterday, because this becomes the foundation everything else depends on, and it matters more as consumption based pricing makes it expensive to get wrong.She is also unusually practical about legacy technology, including the option of virtualising homegrown tools and layering modern capability over them rather than waiting years to replace them.

Season seven of GAEA Talks opens with Graeme Scott and Nathalie Nahai.Season seven runs alongside our Forrester season and takes a different shape. No agenda, no deck, no running order. Graeme and Nathalie last sat down a year ago, and this is the two of them picking it back up to work out what has changed since, and where it is going, from a human point of view. Nathalie is a behavioural scientist and author working on the psychology of technology, persuasion and influence. Neither of them knew what the conversation would be about when it started.It begins with what Nathalie has been hearing across the rooms she has been in, and it is not what you would expect. Across sectors, and across notably individualistic cultures, the theme coming up again and again is collective agency. How decisions get made together rather than alone. She describes a salon at Harvard's Center for Human Flourishing where roughly five percent of the room felt they had any power to shape how AI develops or gets regulated. The questions are being asked anyway, and unlikely alliances are forming across political divides.Much of what follows is about language. The genie, the magic, the fairy dust. Nathalie points out that mythos originally meant a story that was not literally true but illuminated something deeper, and that what we have now is the inverse. She objects to the word hallucination because it implies a consciousness that is not there. Graeme's half of it is that the magical language worked like a mind trick, bypassing every stage of procurement, security by design and basic controls any other technology would have had to clear. Most of what is now argued about as guardrails would not be needed if the basics had been done once.The sharpest moment is Nathalie on accountability. If a person came into your house, took your data, used your cards and sold your details on, they would go to prison, because accountability exists in the body. There is no equivalent here, and that absence explains a great deal.They go a long way off the map from there. Honeypots against a billion phones. Why homogenised language narrows what can be thought. The World Cup boots that all came out the same colour because everyone was working from the same data. Graeme's art college, eleven floors of students all desperate to be distinctive and all identical within their year. Surveillance and the chilling effect. Where autonomous decision making is genuinely new, and where context makes any single rule useless.It closes on Seligman's learned helplessness experiments. Dogs given unpredictable shocks in cages, and then, when the cages were opened and treats put down, many would not walk out. That, Nathalie says, is the one outcome to avoid at all costs, and the evidence of people choosing otherwise is already there if you look for it.

Graeme Scott sits down with Rowan Curran, Principal Analyst at Forrester, who leads the firm's research on generative AI technologies, tools and strategy, and covers AI platforms, cognitive search and synthetic data. He pioneered Forrester's predictive analytics research in 2015, left for a spell in the public sector because he wanted to sit with problems rather than move on to the next client, and returned in 2022 covering both cognitive search and AI platforms. ChatGPT launched weeks later, putting him at the exact intersection of the technologies that mattered. He calls it the smartest accidental thing his brain has ever done.The first half is the most useful account of agentic AI testing you will hear, and his assessment is unsparing. He describes testing as one of the most horrifyingly limited areas of the entire agentic space. Model evaluations and leaderboard scores are directionally useful and make good marketing, but they do not translate into enterprise outcomes. Most organisations have no evaluation set for the task they are automating, so they cannot say whether the system works. Manual testing does not scale to an enterprise full of agents. Using a model to judge another model means first building trust in the judge, which he describes as turtles all the way down.He has been asking agentic platform vendors about their testing suites for two and a half years. Through the end of 2024 and most of 2025 the answer was that it was a next quarter problem. Only in the first half of 2026 has he seen anything he would call robust.The second half widens considerably. He argues the existential risk framing obscures what these mostly are, which is software and cybersecurity risks. Humans will not be destroyed by a tool. Humans will do it with one, negligently or deliberately. He is equally clear that any regulatory framework has to be designed so it does not hand control to the model companies, and draws the parallel with letting social media platforms set their own age limits.He is also sharp on what the field has forgotten. The algorithmic bias work of the late 2010s, facial recognition failing on darker skin tones, the accessibility opportunities in language and vision models, has largely dropped out of the conversation, and that loss shows up in what gets built.Then the line that gives the episode its sting. The marketing goal of creating an audience of one is an extremely lonely and sad place for every human to be in. If your content and mine share no source, what are we going to talk about? That leads into an exchange on music, memory and what a shared cultural soundtrack was for, including what happened when Graeme generated northern Iraqi folk music and played it to people who grew up with the real thing.