consciumai
Conscium’s mission is to help develop safe and efficient AI
Machines may acquire consciousness in the next decade or so. We need to be prepared.
AI agents are coming. They need to be routinely verified.
- Robots are improving fast. This biomimetic one from China will cost $173,000. It is still well inside the uncanny valley, but for how much longer? www.youtube.com/watch?v=jtyE...
- A Chinese robot named Bolt outruns humans. www.youtube.com/watch?v=rucN...
- "Moltbook is the six-finger phase of agent interaction." Nice line from the Hard Fork podcast.
- New research compared leading AI models and 100,000 human using a standardized psychological test for creativity. The best LLMs are now more creative than the average human, but less than the most creative humans. singularityhub.com/2026/01/27/a...
- Big news for Conscium: Our CEO Daniel Hulme is one of the founding fellows of the Academy for the Mathematical Sciences, alongside luminaries like Davids Silver and Spiegelhalter. www.acadmathsci.org.uk/2026/01/29/a...
- Stanford economist Charles “Chad” Jones: "the effect of AI will be perhaps more than 10x the internet... It would be prudent to [prepare for the consequences for labour markets, inequality, and catastrophic risk.” Yup. H/T Anthropic's Jack Clark
- 6/6 Forced to choose whether the first superintelligence is conscious or a zombie, Chris would choose conscious for the intriguing reason that we would probably evolve by uploading, and the mind we upload to join should be conscious. www.prism-global.com/podcast/chri...
- Shane Legg, one of the three founders of DeepMind, has always been upfront about superintelligence being near. He is now hiring an economist to solve the Economic Singularity. The other Big AI firms should do the same.
- 5/6 Chris speculates that mind crime may not be a major problem if the AI which is conscious is a superintelligence, because it could edit its own mind to relieve any suffering caused by human error or bad luck. www.prism-global.com/podcast/chri...
- Conscium on "The Business of Thinking" podcast. AI agents, verification, and the Economic Singularity. And a discussion about whether making machines conscious could actually make them safer. www.youtube.com/watch?v=ilBf...
- Joscha Bach's CIMC publishes "The Machine Consciousness Hypothesis" , an argument for computational functionalism. Oddly, no mention of CIMC's claim that the first superintelligence should be conscious, not a zombie. cimc.ai
- Anthropic’s latest "Constitution" for its Claude speculates that it may be conscious. “We believe that the moral status of AI models is a serious question. ... some of the most eminent philosophers on the theory of mind take this question very seriously.”
- Glimmers of hints that AI is starting to boost productivity. Bubble, what bubble? weightythoughts.com/p/ai-gains-s...
- 3/6 Chris is exploring how well-informed but non-expert people change their opinions about artificial consciousness when exposed to evidence from philosophy, social sciences, and lab experiments. www.prism-global.com/podcast/chri...
- 2/6 Experts on consciousness and artificial consciousness don’t seem to be converging on a consensus. Functionalism is the largest group, but previously fringe positions like panpsychism and illusionism are resurgent. www.prism-global.com/podcast/chri...
- 1/6 Our guest this episode is Chris Percy, Director of the Co-Sentience Institute. He is working to map and improve the resources that will enable people to make informed decisions about artificial consciousness. www.prism-global.com/podcast/chri...
- McKinsey CEO Bob Sternfels says the firm's 65,000 employees now include 25,000 AI agents. This is not the start of the economic singularity, though. Since 2023 the firm has reduced its human headcount by 5,000, following pandemic over-hiring, and added the AI agents.
- The good news. London's mayor is beginning to think about the economic singularity. The bad news. He thinks it can be deferred. news.sky.com/story/sadiq-...
- Microsoft's AI Economy Institute published this map of AI adoption globally. The USA comes 24th.
- CNBC asked 40 business and tech leaders if there is an AI bubble that is about to burst. This chart shows their answers. Which, average to "meh". Details at: www.cnbc.com/2026/01/10/a...
- Ten years after leaving the Human Brain Project, Henry Markram is back in the brain business. www.openbraininstitute.org/about
- This episode is different: no guest. David and I Iook back at 2025 and forward to 2026. We discuss the rapid improvement of LLMs, the AI bubble, AI agents, artificial consciousness, and the future of jobs. calumchace.com/podcast-feed/
- This episode is different: no guest. David and I Iook back at 2025 and forward to 2026. We discuss the rapid improvement of LLMs, the AI bubble, AI agents, artificial consciousness, and the future of jobs. calumchace.com/podcast-feed/
- This episode is different: no guest. David and I Iook back at 2005 and forward to 2006. We discuss the rapid improvement of LLMs, the AI bubble, AI agents, artificial consciousness, and the future of jobs. calumchace.com/podcast-feed/
- The best list of forecasts for AI in 2026 I've seen so far. www.understandingai.org/p/17-predict...
- Will 2026 be the year when the phone is replaced as our window to the internet? Probably not. But smart glasses powered by a phone in your pocket are coming. techcrunch.com/2025/12/30/t...
- Tech investors debate whether 2026 is the year the Economic Singularity starts. (Answer: probably not. Automation will accelerate, causing rapid job churn, but not more lasting unemployment.) techcrunch.com/2025/12/31/i...
- 13/13 Stephen has an intriguing vision for a future version of his book, which evolves according to each reader’s requirements. He would train “WittGPT” on his voluminous notes, and it would write the ideal book for each reader. calumchace.com/podcast-feed/
- 12/13 Huang’s big idea for the next five years is robots. Nvidia is investing massively in physical AI, which he is developing in simulated environments. calumchace.com/podcast-feed/
- 11/13 Jensen Huang is in denial about the economic singularity. This is a problem. calumchace.com/podcast-feed/
- 10/13 Nvidia’s huge stock market success has enabled it to offer staff a very generous share option programme. More than half of its 40,000 people are multimillionaires. calumchace.com/podcast-feed/
- 9/13 Nvidia’s customers would love to have a cheaper alternative supplier. One candidate is Google’s Tensor Processing Units, or TPUs, which are more efficient than Nvidia’s GPUs for the matrix math processing required in AI training and inference. calumchace.com/podcast-feed/
- 8/13 The CUDA software platform has been optimized for GPUs for two decades. It has millions of lines of code, a massive developer community, and a huge library of tools and applications. It is not easy to use, but has become a powerful moat for Nvidia. calumchace.com/podcast-feed/
- The jagged edge of AI.
- Robots are now better than humans at dancing like robots. Which only seems fair. www.youtube.com/watch?v=52cM...
- 7/13 Huang noticed that scientists were hacking Nvidia’s graphics cards to make them into low budget supercomputers. He decided to give them the tools himself, and this spawned CUDA (Compute Unified Device Architecture), released in 2006. calumchace.com/podcast-feed/
- 6/13 Nvidia began as a supplier to the gaming industry, but it pivoted towards AI when the AI big bang of 2012 happened, with the arrival of deep learning. Nvidia now supplies around 90% of GPUs for the AI industry. calumchace.com/podcast-feed/
- 5/13 When Nvidia was still producing graphics cards for video games, it was competing against 70-odd companies. He adopted a strategy called “brain extraction”, which meant hiring the top scientists away from his top rivals, which collapsed them. calumchace.com/podcast-feed/
- 4/13 Huang is a "single-minded and ferocious leader" and Witt describes his intense, even humiliating, questioning style. His so-called "light-speed" management philosophy means that he is relentless in seeking ways to accelerate business processes. calumchace.com/podcast-feed/
- 3/13 Huang was sent to a Kentucky boarding school his family thought was an elite institution, but was actually a school for juvenile delinquents! Despite speaking little English, Huang survived and even thrived here, and acquired “street smarts”. calumchace.com/podcast-feed/
- 2/13 Jensen Huang was born into a middle class family in Taiwan in 1963, and his family moved to the US when he was a child. Sadly, his name does not come from the iconic car brand, but is an Anglicization of his Chinese name, Jen-Hsun. calumchace.com/podcast-feed/
- 1/13 Our latest guest is Stephen Witt, an American journalist and author. His latest book, "The Thinking Machine”, is about Jensen Huang and Nvidia, and has won the 2025 Financial Times and Schroders Business Book of the Year Award. calumchace.com/podcast-feed/
- Can consciousness make AI safer? Conscium CEO Daniel Hulme explores the pros and cons of zombie vs aware superintelligence. conscium.com/can-consciou...
- 6/6 If Cameron had to choose whether the first superintelligent AI is conscious or a zombie, he would choose a conscious one. www.prism-global.com/podcast/came...
- Robots are getting pretty impressive. www.youtube.com/watch?v=rRCX...
- 5/6 This work has led Cameron to become less sceptical than he was about LLMs being conscious. His P(Consciousness) has climbed from 1% to as high as 20%. So he worries a lot about Mind Crime. www.prism-global.com/podcast/came...
- 4/6 Cameron also found that when LLMs are prompted to be self-referential, they seem to become more alike than otherwise. I.e., Gemini, Claude and ChatGPT sound more like each other. www.prism-global.com/podcast/came...
- 3/6 We discuss how the idea that we only use 10% of our brains, the premise of the movie Lucy, is true in a literal sense, but wrong for all practical purposes. www.prism-global.com/podcast/came...
- 2/6 Cameron’s most recent paper noted that self-referential processing by LLMs yields reports of subjective experience - especially when their deception capabilities were suppressed. www.prism-global.com/podcast/came...
