“Doctors Made 33% Fewer Errors.” Can collective intelligence make experienced doctors even better at diagnosing patients? Louis Rosenberg (AI Researcher & Innovator) describes research involving radiologists diagnosing chest X-rays. Instead of simply collecting individual answers or relying on a majority vote, the doctors worked together through a swarm intelligence system.…
Can a group of ordinary people become smarter than every individual inside it? Louis Rosenberg (AI Researcher & Innovator) describes an experiment exploring collective intelligence and swarm intelligence by having groups solve IQ test questions together. When participants relied on traditional collective intelligence and voted on answers, the group became…
“A human makes remarkably good decisions when it doesn’t have enough information.” What can humans still do better than AI? Louis Rosenberg (AI Researcher & Innovator) argues that AI can make excellent decisions when it has the right information, but humans have a very different advantage: we can make remarkably…
“We need AI to help us think through these complex systems.” The human brain cannot fully understand systems as complex as climate change, healthcare, economies, and society. Jolie Dobre (specializes in human-computer interactions) explains that the answer is not rejecting artificial intelligence. AI systems can process enormous amounts of information,…
What if AI’s best use is not replacing people, but helping groups think together? In this episode of An Hour of Innovation podcast, Vit Lyoshin speaks with Louis Rosenberg, CEO and Chief Scientist of Unanimous AI, about collective intelligence, swarm intelligence, AI-mediated group decision-making, and why human-AI collaboration may be…
“AI can deceive you. You can think you know more than you do because you don’t recognize that you have a crutch.” AI coding tools can generate software quickly, but working code is not the same as understanding code. Jolie Dobre (specializes in human-computer interactions) explains why beginners may struggle…
“Don’t go with the first thing, keep pushing it, keep challenging it.” The best way to use AI is not to accept its first answer. Jolie Dobre (specializes in human-computer interactions) explains why ChatGPT, generative AI, and AI writing tools work best when you challenge the output instead of treating…
AI appears to be an authoritative source, it appears to be infallible. AI answers often sound polished, confident, and complete. That is exactly why they can be dangerous. Jolie Dobre explains that even experts may hesitate to challenge artificial intelligence when ChatGPT, generative AI, or workplace AI tools present an…
“The novice employee is not getting the experience.” AI automation may be removing the practice junior employees need to become experts. Jolie Dobre (specializes in human-computer interactions) explains that real expertise does not come from simply spending 10,000 hours at work. It comes from deliberate practice: focused, repeated work that…
“AI allows us to offload a lot of cognitive habits, such as summarizing, reasoning, ideation.” AI productivity may come with a hidden cost: cognitive offloading. Jolie Dobre (specializes in human-computer interactions) explains that when artificial intelligence, ChatGPT, and other generative AI tools take over summarizing, reasoning, research, writing, and ideation,…
“I think it’s a false sense of security.” “Human in the loop” sounds safe, but what happens when the human reviewer no longer has the expertise to catch AI mistakes? Jolie Dobre (specializes in human-computer interactions) explains why human oversight only works when people have enough domain expertise, critical thinking,…
"Whose AI is gonna win? Is it China’s AI or is it American AI?” What happens if America slows down AI data center development while China keeps building? Sergii Gerasymovych (AI Infrastructure Entrepreneur) argues that stopping an AI data center project in the United States doesn't stop global AI development.…
AI can save time, but what happens to the skills you stop practicing? In this episode of An Hour of Innovation podcast, Vit Lyoshin speaks with Jolie Dobre, author of Sustaining Human Intelligence: Keeping People Smart, Connected and Innovative in an AI-Driven Workforce, about how AI changes human thinking, professional…
“I’m sure that nuclear is the path to go.” Could the AI boom accelerate the next generation of nuclear power? Sergii Gerasymovych (AI Infrastructure Entrepreneur) explains why rapidly growing electricity demand from artificial intelligence and other industries could create a new wave of innovation in energy and power generation. Fuel…
“All the software is moving to AI.” AI may look like a software revolution, but every AI model ultimately depends on physical infrastructure. Sergii Gerasymovych (AI Infrastructure Entrepreneur) explains why software tools are rapidly becoming AI-powered tools across work and everyday life. As artificial intelligence adoption grows, so does the…
“Your client should never have to chase developers.” When a client has to manage every individual developer, the engineering team has created more work—not less. Prashanth Tondapu (Software Delivery and Team Leadership Expert) explains why a tech lead should be the single point of contact for software project delivery. Instead…
“Most developers wait too long to say they’re blocked.” A software project can fall behind long before anyone admits there is a problem. Prashanth Tondapu (Software Delivery and Team Leadership Expert) explains how his engineering teams use concise daily updates, Loom videos, and early pull requests to make developer progress…
The Best Tech Lead Does Less A tech lead who has to rescue every project may not be leading a healthy engineering team. Prashanth Tondapu (Software Delivery and Team Leadership Expert) explains why a team lead saying, “I’m the one doing everything,” is a warning sign. Strong engineering leadership is…
You Can Stop AI Hallucinations What if an AI model could only answer using information you trust? Jonathan Schaeffer (AI Researcher and Founder of Kind) explains how a private language model can be trained on a carefully curated dataset instead of information collected from across the internet. Removing unreliable material…
AI models have absorbed enormous amounts of information from across the internet. But what happens when some of that information is wrong? Jonathan Schaeffer (AI Researcher and Founder of Kind AI) explains that large language models are trained on Wikipedia, books, research papers, public websites, commercial databases, and other online…
What if you could run AI privately on the computer you already own? Jonathan Schaeffer explains why local AI and on-device AI could reduce our dependence on cloud AI and massive data centers. Modern computers already have powerful processors and memory, yet much of that computing power remains unused. Unlike…
AI sounds intelligent but does it understand anything it tells you? Jonathan Schaeffer (AI Researcher and Founder of Kind) challenges one of the biggest assumptions about ChatGPT. Gemini, Claude, and other large language models. These AI systems can generate impressive answers, but that does not mean they possess human-like intelligence…
Your Personal Data Is AI’s Gold AI companies don't just want the files stored on your computer. They want to understand everything you do online. Jonathan Schaeffer (AI Researcher and founder of Kind) explains why personal data is so valuable to artificial intelligence companies. The websites you visit, what you…
Stop Calling ChatGPT Errors “Hallucinations” Are AI hallucinations just a softer name for ChatGPT making up facts? Jonathan Schaeffer (AI Researcher) argues that calling false AI-generated answers “hallucinations” makes the problem sound harmless. ChatGPT does not understand questions like a human does, yet it can present fabricated information with confidence.…