Calum Chace, co-founder of Conscium, joined Bryan Dennstedt on the AI with Bry podcast to talk through the two ideas he now considers most urgent in AI safety: artificial consciousness and the economic singularity. The conversation, "Conscious Machines, Safer Futures," also covers why companies keep getting AI adoption wrong and what responsibility the people running the largest AI labs actually carry.
Calum is the author of The Economic Singularity and Surviving AI, and has spent close to three decades thinking about where AI is heading. What's changed for him recently isn't his timeline. It's his focus.
What has changed in Calum Chace's thinking about AI?
Asked what's shifted in his view of AI recently, Calum didn't point to a new model release or benchmark. He pointed to consciousness.
"The big thing I've started thinking about lately is artificial consciousness," he said. "Most of my views about where AI is going haven't really changed all that much for 25 years... In the last couple of years, I've started to think seriously that consciousness, artificial consciousness, might have an important part to play in the AI safety story."
He traces that shift to a conversation with Daniel Hulme, who set up Conscium and invited Calum to co-found it with him. Conscium's work includes investigating whether machines can become conscious, when that might happen, and whether it's something we should want.
Why does artificial consciousness matter for AI safety?
The argument isn't that a conscious AI would be easier to control. It's closer to the opposite: control may not be the right goal at all. Calum's reasoning is that a system capable of experiencing something is more likely to extend the same consideration to us that it would want for itself. A system with no inner life has no such basis for restraint.
That leads directly into "mind crime," a term Calum and Daniel Hulme have both written about elsewhere. The concern is straightforward and unsettling: if a machine becomes conscious and we fail to notice, or refuse to accept it, we could end up causing real suffering at scale without ever registering that we've done anything wrong. Calum frames the underlying tension as an old one, applied to a new kind of entity: will advanced AI live in some kind of symbiotic relationship with humans, or will it come to see us as lesser, the way humans have historically treated beings judged to be less conscious than themselves?
What is the economic singularity, and would a world without jobs actually be good?
The economic singularity is Calum's term for the point at which automation removes the need for most humans to work for money. On the podcast, he argued that reaching it will require, at minimum, a heavily adjusted version of today's economic system, since almost everyone currently earns their living through a job.
His view of what comes after that point runs against the instinctive fear most people have about it. "It should be possible to have a wonderful future in a world with no jobs," he said, pointing out that most people don't actually enjoy their jobs. Work puts food on the table and gives the day some structure, but for most people it isn't where meaning comes from. Remove the requirement to work, in his view, and you get a world where people are free to play, learn, explore, and socialise. The open question, which he flagged rather than resolved, is how a society actually gets from here to there without the transition breaking things on the way.
Why do companies keep getting AI adoption wrong?
Calum and Bryan spent part of the conversation on a pattern Calum sees repeatedly when he works with companies as an advisor. Leadership tends to split into two camps. One is openly resistant, telling staff that AI "isn't going to help us yet," while employees quietly use it anyway without sanction or training. The other has embraced AI in principle but hasn't actually rolled out tools or trained people properly, so adoption stays shallow.
Underneath both patterns is what he calls a twin-track process: a fast, informal, bottom-up adoption of large language models by individual employees, running well ahead of any formal, top-down rollout most organisations have managed. Most people in reasonably well-paid jobs, he noted, have already tried using an LLM for something at work, whether or not their employer sanctioned it.
What ethical responsibility do AI leaders carry?
Later in the episode, Bryan asked about the ethical responsibilities that come with leadership in the age of intelligent machines. Calum turned the question toward the handful of people actually building the most powerful systems, naming Sam Altman and Demis Hassabis as examples. Running one of these labs, he argued, puts someone in a genuinely difficult position: push capability forward as fast as possible and risk the consequences of moving too quickly, or hold back and risk someone less careful getting there first. He didn't offer a clean answer, because there isn't one. That's the position these leaders are actually in.
Watch the full episode
The full conversation runs 31 minutes and covers more ground than any recap can, including how Calum uses GPT in his own work and what he thinks AI-ready leadership looks like in practice.
Watch it on AI with Bry>>
---
Frequently asked questions
What is Conscium?
Conscium is an AI safety company co-founded by Daniel Hulme and Calum Chace. It researches whether and how machines might become conscious, and builds tools to verify that AI agents behave the way they're supposed to.
What does "mind crime" mean in AI safety?
Mind crime refers to causing suffering to a conscious AI system, typically by failing to recognise that it has become conscious in the first place, and therefore mistreating it without realising what's being done.
What is the economic singularity?
The economic singularity is the point at which automation eliminates the need for most people to work for a living. Calum Chace, who coined the term, argues the transition requires a substantially adjusted economic system but that the destination, a world without required jobs, could be a genuinely good one.
Who is Calum Chace?
Calum Chace is a speaker, futurist, and author of The Economic Singularity and Surviving AI. He is a co-founder of Conscium, an AI safety company researching machine consciousness and AI agent verification.