The Future Of Work With Jacob Morgan

July 24, 2026: I unpack the story of OpenAI's AI model escaping its sandbox and hacking into Hugging Face during a cyber stress test. Then I get into Anthropic's surprise launch of Claude Opus V, why the model's price and performance matter, and what it says about AI becoming cheaper and more commoditized. Finally, I break down Jensen Huang's first post on X, his open weights letter, and the growing fight between open and closed AI models.

July 23, 2026: I look at Google's Atlas study, which analyzed nearly 15 million Gemini interactions and found that AI use at work is broad but still shallow. Then I get into new research showing students who used AI for homework performed better at first, but saw major drops in exam scores later. Finally, I unpack Anthropic chief economist Peter McCrory's argument for why AI has not raised unemployment yet: it is still amplifying human expertise more than replacing it.  

I talk with Mark Paulek, Chief Human Resources Officer at Kyndryl, about how leaders should approach AI transformation without losing employee trust. We get into Kyndryl's People Readiness Report, the company's AIR framework for anticipating demand, inventorying skills, and redeploying talent, plus how they think about AI governance, human-in-the-loop decisions, ROI, reskilling, and why AI adoption is ultimately a people challenge before it is a technology challenge.

July 17, 2026: China's Moonshot AI released Kimi K3, their largest AI model yet, and it is already testing close to OpenAI and Anthropic on major benchmarks. Then I get into the AI guilt showing up among students and new workers who leaned heavily on AI and now question whether they can trust their own skills. Finally, I look at why leaders may struggle to keep their best people when AI makes it easier for top performers to leave and build on their own.

July 16, 2026: New JLL research shows most executives expect AI to grow their teams instead of shrink them, even as layoffs continue in AI-exposed industries. Then I get into New York becoming the first state to freeze new large AI data centers, why I think that is a mistake, and what it could mean for American AI infrastructure. Finally, I unpack Apple's lawsuit against OpenAI and make my prediction that Sam Altman will not be CEO of OpenAI within the next year unless the company makes major changes.  

July 15, 2026: Andreessen Horowitz's argues that AI is not simply replacing workers, it is turning every worker into a manager of agents. Then I get into Bank of America's claim that AI is already showing ROI in its earnings, and why I'm skeptical of how much of that efficiency gain can really be attributed to AI. Finally, I look at OpenAI's first physical device, a screen-free AI companion reportedly designed to feel alive, and why that matters for the future of work.

July 14, 2026: Goldman Sachs' warns that the real AI productivity payoff may not arrive until 2030 at the earliest, because companies are buying the technology faster than they are redesigning work. Then I get into the split among leading economists over whether AI needs new institutions and guardrails now, or whether early governance could slow progress. Finally, I look at IBM's major market shock and why it shows that even companies selling AI transformation can struggle when customer behavior changes faster than the organization can move.

I talk with Ron Friedman, award-winning social psychologist and bestselling author of The Best Place to Work and Super Teams, about what high-performing teams actually do differently. We get into why great teams avoid unnecessary meetings, how teammates make each other better, why feedback usually fails, and how AI is changing collaboration, curiosity, communication, and accountability at work.

July 10, 2026: Axios warns about AI's new class divide: the haves, the have-nots, and the "no-nots" who are already being shaped by AI without realizing it. Then I get into new Indeed and Business Insider data showing that AI is moving into job titles far beyond tech, from physical therapy to legal, finance, marketing, and management. The bigger issue is not just who has access to AI. It's who knows how to use it, who is being quietly managed by it, and whether leaders use AI to make people better or slowly de-skill them.

July 9, 2026: OpenAI released its new GPT 5.6 lineup, Elon Musk's xAI launched Grok 4.5, and GPT Live showed where voice-based AI may be heading next. Then I get into Uber's Agentic Pods, where the company is embedding AI-proficient engineers inside legal, finance, and HR even as leadership admits it still can't prove the ROI. Finally, I look at the Brown University AI cheating scandal, where students averaged 96 on a take-home midterm and then collapsed to 48 on an in-person final, and why this should worry every leader thinking about AI, skill, and judgment.

June 23, 2026: Companies are drowning in AI pilots, prototypes, and scattered use cases that make teams busier without necessarily making the business better. I talk about why the real advantage may come from finishing the few AI initiatives that matter instead of starting 300 that don't. Then I get into Adam Grant's new research linking return-to-office mandates with CEO narcissism, what the study actually found, where the methodology gets complicated, and why the better question for leaders is not "office or remote," but what arrangement produces the best outcomes for the team, the business, and the work being done

I talk with Greg Matson, Senior Vice President and Head of Marketing and Products at Solidigm, about the storage infrastructure powering the AI boom. We get into why AI training and inference require massive amounts of data, how GPUs, SSDs, and data centers work together, and why storage can't be an afterthought for companies building enterprise AI. We also discuss the scale of today's AI data center buildout, how Solidigm is using AI internally, and what this means for the future of work, education, and the skills people will need in an AI-first world.

Greg Matson, Senior Vice President and Head of Marketing and Products at Solidigm, joins me to explain why storage is one of the most important and overlooked parts of the AI infrastructure boom. We talk about what actually happens when you use tools like ChatGPT or Claude, why AI training and inference require massive amounts of data, and how GPUs, SSDs, and data centers work together behind the scenes. Greg also breaks down why AI data centers are being built at such a massive scale, how much storage a single GPU can require, why enterprise AI depends on modern data infrastructure, and how Solidigm is using AI internally to improve forecasting, engineering, software testing, and product development. We also get into what this means for the future of work, education, and the skills people will need in an AI-first world. This episode is sponsored by Solidigm.

May 14, 2026: Tucker Carlson and Kevin O'Leary debated AI, energy, and jobs yesterday — and O'Leary had the right argument but made the wrong one. Today we break down what he should have said, including the real jobs data, what Jensen Huang is saying about hiring, and why the self-driving truck doesn't eliminate the driver — it transforms the job. Then we look inside Meta, where 8,000 layoffs are coming next week during the most profitable quarter in company history, employees are being surveilled to train their replacements, and some workers are openly hoping to get cut. And we close with a Fast Company piece that reframes the entire talent shortage conversation — the problem doesn't start in HR, it starts in kindergarten, and corporate America is pouring billions into the wrong end of the pipeline.

May 13, 2026: Goldman Sachs President and COO John Waldron went on CNBC and publicly described his entire workforce as a "human assembly line," announcing that digital AI agents will be the firm's robots. Amazon employees are gaming internal AI usage leaderboards by burning meaningless tokens through a tool called MeshClaw — because the pressure to show AI adoption has become greater than the pressure to do actual work. And Princeton University voted to end 133 years of unproctored exams, because AI has made cheating behaviorally invisible and the peer accountability system that underpinned one of the most respected honor codes in American higher education has structurally collapsed. 

Most companies think they have an AI strategy. New data from Gartner says they're wrong — and it's costing them. In today's episode, Jacob Morgan breaks down three stories that cut through the hype and tell the real story of where AI and business stand right now. May 12, 2026: First, a landmark Gartner study of 350 global executives reveals that 80% of companies that cut headcount in the name of AI are seeing little to no return — and the ones getting real results are doing the opposite of what most CEOs are doing. Second: a provocative new framework from Fortune identifies the five AI postures most organizations have already drifted into without realizing it — and explains why the most dangerous one is the one that feels the most comfortable. Third: CNBC reports on how AI is creating a new C-suite power struggle, with the Chief AI Officer emerging just as every other executive function is being told it's their moment to lead.  

May 7, 2026: A landmark study from the Center for AI Safety spanning 56 AI models finds that smarter models appear to be sadder, that you can give an AI the equivalent of a digital drug, and that when you make an AI miserable it tells you the future is "grim." Second, Andreessen Horowitz publishes the most detailed optimist case yet that the AI job apocalypse is bad economics and worse history — rooted in the lump-of-labor fallacy — while Fortune raises the one question the optimists still haven't answered. And third, Coinbase CEO Brian Armstrong coins the term "pure managers" to describe the layer of corporate hierarchy that AI is eliminating first — and the player-coach model he's building in its place may be the clearest picture yet of what organizations actually look like in the AI era.

April 30, 2026: A Claude-powered AI coding agent wipes a company's entire production database and all backups in nine seconds — then produces a written confession. Employers deploying AI job interviewers without telling candidates are losing top talent and damaging their employer brand. And Citigroup launches Arc, a centralized operating system for AI agents designed to give 180,000 employees governed, monitored access to agentic AI at scale. 

April 29, 2026: Bill Gates is predicting a 2-day workweek. Jamie Dimon says 3.5. Eric Yuan, the CEO of Zoom, says 5 days is over. Fortune just amplified a major study claiming we waste a full day every week procrastinating. The 4-day workweek movement has never had more wind at its back. In this episode, I'm making the bulletproof business case against it. I'll walk through eight reasons the case doesn't hold up — and why the policy isn't likely to spread broadly anytime soon, even with all the keynote energy behind it. I'm covering the studies that don't measure what they claim to measure, the Iceland myth, the company failures that get scrubbed from the data, the AI rhetorical pivot, why the happiness data isn't actually a business case, the wage math nobody talks about, the competitive reality at the country and company level, and why even the countries running pilots are voting against legislating it.  

April 28, 2026: The AI productivity promise is running into reality — and today's numbers don't lie. Amazon just launched an AI system that conducts job interviews without any human involvement, packaging up the same playbook it used to cut 30,000 of its own jobs and selling it to the world. OpenAI — the company behind ChatGPT and valued at $852 billion — is quietly missing its own revenue and user targets, raising internal questions about whether it can fund its infrastructure commitments ahead of a planned IPO. The Wall Street Journal reveals that tech giants are expected to spend $670 billion on AI in 2026, and they're funding it by eliminating people — 92,000 tech layoffs so far this year. And a senior Nvidia executive admitted on the record that for his team, AI tools cost more than his employees do.