\boxed{“`html
Dr. Alan D. Thompson is a globally recognized AI expert and the mind behind The Memo—the world’s most widely read AI newsletter. He recently joined Cris Sheridan on FS Insider to unpack the seismic shifts taking place in robotics and AI. With his finger firmly on the pulse of advancements, Thompson’s insights, relied upon by governments and corporations alike, illuminate a world where humanoid robots are stepping into factories and homes, large language models (LLMs) are racing toward superintelligence, and the very fabric of human labor and economics is being rewoven. In a conversation brimming with revelations, Thompson shared his excitement—and his predictions—about a future that’s arriving faster than anyone anticipated, especially with his AGI (artificial general intelligence) countdown now at 92%.
Audio discussion: 90% to AGI: How AI and Robotics Are Reshaping Global Markets in 2025.
Humanoid Robots: From Factory Floors to Kitchen Counters
The past few weeks have marked a turning point for AI, and for Thompson, the spotlight shines brightest on humanoid robots. “It’s almost like someone’s lit a match under 14-16 of these humanoid robot companies,” he remarked, capturing the explosive progress in a field that’s been simmering for years. These aren’t just clunky machines—they’re sophisticated systems designed to look and act like humans, now hitting production lines at giants like Foxconn and Audi, and even poised to enter our homes.
Thompson highlighted a global race, with standout players emerging from unexpected corners. Norway’s 1X Neo Gamma, a fully clothed robot capable of peeling plastic off a new phone or sorting a wardrobe, tops his list of favorites. China’s Unitree H1 and UBTECH Walker S1—powered by the latest Chinese DeepSeek R1 model—are making waves in factories like Volkswagen and BYD. Meanwhile, Figure’s 02 and 03 humanoid robots—powered by their Helix AI model—are breaking barriers, available not just in industry but also for domestic use.
What’s driving this surge? A perfect storm of innovation—improved battery chemistry, tactile sensitivity, and plummeting costs. Thompson predicts prices could drop from today’s $30,000–$40,000 range to as low as $10,000–$20,000, or even lower, as robots like Apptronik’s Apollo begin building and repairing themselves. “If everything’s being created, repaired, and maintained [by robots and AI]”, he mused, “there wouldn’t be much sense in having an economic model like we have now.”
The Agentic AI Leap: Autonomy Redefined
Beyond physical robots, Thompson sees another frontier accelerating: agentic AI. These are LLMs that don’t just respond—they act autonomously, tackling complex tasks without pre-scripted routines. “This is not the concept of running a loop or a script,” he explained. “It will encounter something it’s not seen before and give the most likely best response.” From automating insurance workflows in the world’s top companies to replacing teams of analysts, agentic AI is scaling the autonomy ladder at breakneck speed.
Grok definition of agentic AI: “Artificial intelligence systems designed to operate with a high degree of autonomy, enabling them to independently make decisions, execute actions, and work toward predefined objectives. Unlike traditional AI, which often relies on explicit human instructions, agentic AI can adapt to changing circumstances, learn from its environment, and proactively initiate tasks, all while requiring minimal oversight.”
Based on his research, Thompson anticipates a major OpenAI release in this space soon capable of outpacing human teams. The implications? A world where tedious desk jobs vanish, and AI handles everything from document workflows to strategic decision-making—24/7, lights off, no coffee breaks needed.
Superintelligence on the Horizon: 10 Million Einsteins?
Quoting Anthropic CEO Dario Amodei, Thompson painted a vivid picture of what’s at stake: “If someone dropped a new country into the world with 10 million people smarter than any human alive today, you’d ask, what’s their intent?” This isn’t sci-fi—it’s the reality of LLMs reaching genius-level intellect. With 43 new major models released since January 2025 alone, bringing the total to 539, the race is on. OpenAI’s o1 and o3, Anthropic’s Claude 3.7 Sonnet, and China’s DeepSeek R1 lead the pack, excelling on benchmarks like MMLU and GPQA, often outscoring human PhDs.
Background note: MMLU stands for Massive Multitask Language Understanding, a benchmark test that evaluates an AI model’s ability to understand and answer questions across 57 diverse subjects, ranging from high school topics to professional fields like law and medicine. It’s like a broad IQ test for language models, measuring general knowledge and reasoning. GPQA stands for Google-Proof Questions and Answers, which is a graduate-level science test with tough, original questions in fields like biology, physics, and chemistry, designed to be unsearchable online. It assesses an AI’s ability to reason through complex, novel problems, far beyond average human performance (where humans score ~0%).
But it’s not just about smarts—it’s about innovation. Thompson eagerly awaits AI that can invent anew, from adding elements to the periodic table to solving mental health crises. “We’re on the way to having LLMs become innovators,” he said, pointing to his Artificial General Intelligence (AGI) countdown at LifeArchitect.ai/AGI now reaching 92%. Once it reaches 100%, Thompson believes superintelligence (the next step) could come much quicker than many anticipate.
The US-China AI Arms Race
The battle for AI supremacy is intensifying, and Thompson sees it as a San Francisco vs. Beijing showdown currently. “There’s no one in the top 20 outside of the US and China,” he noted, with OpenAI’s models dominating rankings, followed closely by Chinese contenders like DeepSeek’s R1. This rivalry extends beyond software into hardware and governance, with AI now steering military decisions—think Palantir and Scale AI plotting troop movements—and poised to overhaul government services under initiatives like DOGE.
Yet, risks loom. Thompson referenced the emerging MAIM project (Mutually Assured AI Malfunction), a modern echo of the nuclear arms race, as nations grapple with superintelligent systems that may soon defy human control or comprehension.
Looking Ahead
Thompson teased his upcoming Memo, promising deep dives into military AI like NIPR-GPT and Camo-GPT, OpenAI’s $3 billion SoftBank deal, and Sanctuary AI’s touch-sensitive robotic fingers. “It’s an exhausting time to be alive in the AI space,” he admitted, but his mission remains clear: distill the chaos into concise, actionable insights. Readers can catch it at LifeArchitect.ai/memo.
“This is a really exciting time to be alive,” Thompson concluded, reflecting on AI’s ascent to 92% of general human intelligence and capability, based on his research and countdown model. For those tracking the cutting edge, his work remains an indispensable guide to a future that’s no longer distant—it’s here.
For related podcast, see 90% to AGI: How AI and Robotics Are Reshaping Global Markets in 2025. If you’re not already a subscriber to our weekday FS Insider podcast, click here to subscribe.
For a link to our full podcast archive, see Financial Sense Newshour (All) and don’t forget to subscribe on Apple Podcasts, Spotify, or YouTube Podcasts!
To learn more about Financial Sense® Wealth Management, give us a call at (888) 486-3939 or click here to contact us.
Content is for informational purposes only and does not constitute financial, investment, legal, or other advice.
There are risks involved in investing, including the potential for loss of principal.
Forward-looking statements are based on assumptions that may not materialize and are subject to risks and uncertainties.
Any mention of specific securities or investment strategies is not an endorsement or recommendation.
Advisory services offered through Financial Sense® Advisors, Inc., a registered investment adviser. Securities offered through Financial Sense® Securities, Inc., Member FINRA/SIPC. DBA Financial Sense® Wealth Management. Investing involves risk, including the loss of principle. Past performance is not indicative of future results.
“`}
Leave A Comment