
Sign up to save your podcasts
Or


This week’s solo episode tackles arrogance, a trait I know a little too well from my own career. I’ve regretted it every time I acted that way, and I’m seeing the same pattern play out across the AI industry right now.
That includes the Nvidia situation and the Department of Transportation’s aviation grounding, which came three days after an AI system deployment that the FAA has since ruled out as the cause. The administration’s track record with AI rollouts, including the Weather Service, makes these answers worth questioning.
For practitioners, arrogant AI means assuming the technology will work, and that it will act morally, ethically, and effectively without question. In my work with companies and nonprofits, the biggest misses usually come from the same big swing: replacing an entire process, or even a person’s job, without understanding the workflow, the nuance of the task, or where human judgment is needed. When that fails, people decide the technology doesn’t work.
Part of the problem is that we gave it too much power and bought into Silicon Valley’s hype. I see Anthropic’s IPO as a test of how long LLMs can go financially before a crash.
The hope is humility, which means building AI that assists people and takes on rote tasks while keeping humans in their seats. I believe the future belongs to smaller, task-specific models built for fields like science and medicine instead of ever-larger general models like Fable.
I close on leadership and how the arrogance, ethical lapses, and missing succession plan on display at Nvidia show what happens when morals and ethics leave the room. Humility is hard, but it’s worth striving for.
1:20 What arrogant AI means for practitioners
2:32 Why humility matters in AI deployment
2:55 The big swing: replacing whole processes and jobs
4:30 A humbler, assistive approach to AI
4:58 Anthropic’s IPO and the economics of LLMs
6:19 Leadership lessons from the marketplace
This episode is a little different. It’s a dharma-like talk I gave to a small sangha at Radiance Yoga this past Saturday. It was a windy fall morning, with the leaves coming off the trees and change literally in the air. Trees handle impermanence with ease. Most of us, myself included, struggle with it, especially when the change is unwelcome, and AI has become one of the most unwelcome changes in many people’s working lives.
I walk through how I came to Plum Village practice, from a childhood of iron Buddhas and incense to the retreat Caitlin brought me to, and why impermanence stuck with me through the years.. Staying relevant across 30 years of technology change, from fax machines to the web to AI, turned out to have very little to do with technology and everything to do with mindset. I also share how last year’s business collapse and my father’s death two weeks later sent me back to the mat, and how those morning sits became Now Is Gone.
From there, I apply the Four Noble Truths to AI. We suffer when we cling, whether to absolutes like “I will never touch this” or “we’ll cut 30 percent of our workforce,” or to the way our kids, our bodies, and our careers used to be. I can’t stop every software company from bolting on a chatbot or every data center from going up without care for its community.
I can voice my dissent, and I can use the technology where it genuinely helps, like supporting underfunded global health work in India and Africa. My hope is that we pause, reflect, and use AI only where it serves our North Star, like water finding its way to the sea.
Listen today on Spotify, iTunes, YouTube, or wherever you get your podcasts!
1:36 AI as the latest change at work
3:37 Why staying relevant isn’t about technology
4:46 Impermanence and the Four Noble Truths
10:38 Surrender and letting go of what was
11:36 Dukkha: suffering and clinging
16:18 What I can’t stop about AI, and what I can do
18:27 Letting go of absolutes and binary thinking
21:43 Pause, reflect, and respond from your North Star
23:23 Taking stock and discarding what doesn’t serve you
She helped shape the nation’s tech security strategy from the White House. Now she helps companies clean up after sprinkling AI on everything like glitter.
Camille Stewart Gloster has spent her career where emerging technology meets security. She served as the first deputy national cyber director for technology and ecosystem security at the White House, led global product security strategy at Google, and now runs CAS Strategies, a cybersecurity and emerging tech advisory firm. She is also the author of The Insider You Built.
AI is nothing new in cybersecurity, where it has powered anomaly detection for years. Camille’s wake-up call came when ChatGPT launched and her mom called to ask what it was. After leaving government to help companies adopt AI, she found many of them “just kinda sprinkling some AI on things.” Like glitter, it looks fun until you’re the one who has to clean it up.
Her biggest pet peeve is how organizations hand agents the keys to the kingdom. Agents are not people, she cautions, yet they act inside your organization with real access and authority, making decisions on your behalf. We traded intern analogies (plus a hypothetical agent that answers a request for a cheeseburger story by ordering actual cheeseburgers on DoorDash) and landed in the same place: agents need specific instructions, clear boundaries on what they should not do, and human checks.
Camille hopes we land somewhere between Terminator and tech utopia, closer to the utopia, where AI helps solve challenges like climate change and health outcomes and gives us time back to be human. Getting there depends on people making intentional, risk-based decisions about AI, the same way they do about everything else in their lives.
0:49 From the White House and Google to CAS Strategies
2:40 Sprinkling AI like glitter, and who cleans it up
3:51 Building an AI chief of staff without email access
5:14 Why agents are like hiring a new employee
7:26 Smart enough to be dangerous
8:49 Camille’s hope for AI: intentional, risk-based decisions
Learn more about CAS Strategies at https://www.cas-strategies.com/about
Get the Insider You Built: https://www.amazon.com/clp/1394430892
Data centers have dominated the AI news cycle lately, from local concerns about noise and heat to questions about who pays for the added strain on the electric grid. In this solo episode, I take stock of the build-out and why I think it’s happening with very little mindfulness. A conversation last week likened it to the railroad boom of the Industrial Revolution, when tracks went down everywhere, towns boomed or withered depending on where the line ran, and nobody planned for the environmental damage.
Two questions rarely come up in the coverage. First, how much of this capacity is actually needed? Every frontier lab is building its own, with no real effort to treat compute as a shared resource that could lower the environmental impact. Second, why are we running every query through massive general-purpose language models? I compare them to department stores, where everything is on the shelf and none of it is as good as it could be.
Running makes a good example. The AI coaching features in Strava, Runna, and Garmin lean on frontier models, and as a coach I think they suffer for it. A narrow model trained only on running knowledge would likely coach better and draw far less from the grid.
So where’s the hope? I’d like to see companies build data centers alongside environmental groups and communities, pay for their own grid impact instead of passing it to taxpayers, and invest in narrower, more efficient models. For the rest of us, it comes down to mindfulness. Before reaching for a chatbot, ask whether the use case is warranted or just a dopamine hit. Is the scratch really worth it?
0:58 Environmental and grid concerns
2:22 How much capacity do we really need?
4:11 AI as a department store
6:12 Where the hope lies
Last week I gave efficacy versus efficiency about a minute. This week it gets the whole episode. On Thursday I spoke at the GAIN Conference to business-to-government marketers, a room full of contractors and professional services firms still recovering from DOGE while facing heavy pressure to adopt AI. The distinction landed immediately, and it’s resonating with the nonprofits and professional services firms I work with, too.
Efficiency has become AI code for cut costs, cut people, cut, cut, cut. Employees know it, and few employers have promised to bring them along if AI succeeds. Meanwhile, this year’s labor data suggests the promised productivity gains largely haven’t shown up. That leaves a hard question many executive teams can’t answer: beyond doing things cheaper, how does AI help you reach your business goals?
Efficacy is where the hope is. It’s the mission, the North Star, and the brand promise to the customers you serve. If AI helps me reach more people, or improves the quality of what I deliver to the people I already reach, that’s efficacy. It also gives leaders a simple filter for use cases: is this on mission, and does it make us more effective? So what’s your North Star, and which AI use cases actually move you toward it?
0:36 The GAIN Conference and the federal market’s AI moment
2:23 Efficiency as code for cut, cut, cut
5:30 Efficacy as the mission and North Star
7:31 Choosing AI use cases that stay on mission
This solo episode of AI Hope starts with a question I’ve been chewing on for months. AI brings a long list of real problems, from job losses and data center sprawl to cybersecurity, rogue agents, and what I call the dumbification of society. Hollywood and the press amplify the existential fears because doom sells tickets and gets clicks. Meanwhile, industry hasn’t offered a real answer, and few of us expect the federal government to act on our behalf.
My answer is humans. I can’t control what Washington does, but I can control how AI gets adopted in my own organization, and so can you. That starts with use cases that serve your mission instead of reckless, use-it-for-everything deployment.
It also means knowing the difference between efficacy, which is achieving your mission, and efficiency. When I hear executives tout AI efficiencies, I increasingly hear code for layoffs, and unless retraining and redeploying people is part of the plan, you should see yellow and red flags.
I also get into the AI lobby, why elected officials only move when they’re in hot water, and why every agent needs built-in human triggers instead of rogue internet access. Even though I work in this business, I think the frontier model bubble needs deflating. These models are a department store: You get a little of everything, and none of it is great. I expect niche models and SaaS platforms to gain ground once the big IPOs pass.
AI’s biggest problem is that we’ve let it go unchecked, deploying it with a shiny-object attitude that it would cure the world. The fix is us, whether that’s a human check inside a workflow or a bigger strategic call about how your organization uses the technology.
Listen today on Spotify, iTunes, YouTube, or wherever you get your podcasts!
0:42 Is there one answer to all of AI’s problems?
2:23 The answer is…
6:37 The department store problem and the rise of niche models
This week's AI Hope is another solo episode, part of an experiment to see how these commentaries compare to guest interviews. The topic is AI for good, which feels like a radical idea given the week's news. I open with quick reacts to ChatGPT's new agents and the press conference that rebranded AI as super intelligence, a label that still strikes me as ironic for anyone who uses these tools every day.
So can AI be used for good? Yes. I'm working with two nonprofits right now, one focused on the environment and one on global health, and both see AI as a force multiplier for mission impact in a sector that never has enough resources. The politicized "it's evil, we won't use it" stance misses that. Much of the environmental cost comes from oversized large language models most professions don't need, and many of today's "agents" are really automations with probabilistic decision logic.
Is AI evil? No. The danger comes from big tech companies and the government deploying it without rails while offering platitudes about responsibility. That's where we need to keep our eye on the ball. Let me know whether you prefer this monologue format or the guest interviews.
I also touch on the industry's new moral responsibility agreement, which reads like more bluster from companies focused on building value for their investors, and on Anthropic's standoff with the Department of War.
0:36 The superintelligence rebrand
3:17 Can AI be used for good?
8:15 Who is really dangerous?
Amy Kluber and her team at GovCIO cover how federal agencies adopt technology including AI. We recorded this conversation on September 22, the day after AI came up at the UN, capping six weeks of news that ran from security incidents to superintelligence talk to government regulation. As to those comments at the UN, Amy is reserving judgment on superintelligence until she sees AI operate at that level for her.
Personally, Amy’s path into AI was gradual. As federal agencies started prioritizing it, she decided she couldn't judge the technology without using it herself. Now she's working through what AI means inside a newsroom: Whether you can trust its output, if your audience will trust you for using it, and what policies you need.
Her most useful build so far is a Copilot agent that pulls four years of Flywheel Awards nominations out of scattered forms, Word docs, spreadsheets, and SharePoint folders, turning them into a database her team can ask questions of. Her starting point was the right one: find the annoying, repetitive process first, then apply automation to it.
Amy's AI hope is that it shrinks the distance between people and what they're capable of doing, with human judgment still in the loop. For government, that means less time on hold, faster access to services and benefits, and quicker paths to a human when it matters.
1:09 Six weeks of AI news in a nutshell
2:25 AI in the newsroom: trust, ethics, and new policies
3:23 Why her adoption was gradual
5:14 Curing writer's block
5:34 Building a Copilot agent for the Flywheel Awards
7:17 Automating the ankle biters
8:30 How federal agencies are applying AI
8:52 Amy's AI hope
10:02 What AI could do for the American citizen
This week's AI Hope is a solo episode. With everything that's happened, claims of civilization ending AI, cybersecurity failures, data center fights, job displacement, AI slop, IPO hype, is regulation of the AI industry actually coming? My answer is not on the federal level, not for at least the next two years.
Congress isn't functional right now, and big tech is lobbying hard, with a lot of that influence flowing straight to the administration. The administration's own instincts lean toward supporting corporations over individual freedoms, whether that's LLMs claiming the right to scrape anyone's content to train, or defense-related rules aimed at companies like Anthropic.
The real checks on overreach right now are happening on the state and local level and in the courts instead. More than fifteen hundred bills to regulate AI have been introduced, with serious legislative efforts underway in California and Texas, and local fights breaking out over data centers. And individuals suing frontier model companies directly, the case that reined in Meta's algorithmic harm to teenagers, the authors' case I was part of against Anthropic, are proving to be the most effective check we have.
Abroad, the EU now requires AI content creators to disclose or watermark AI-generated material, and Anthropic complied immediately.
Watch or listen to this episode on YouTube! Don't forget to subscribe!
. I don't think meaningful federal regulation arrives in the next couple of years. But voters are angry enough that politicians on both sides are starting to notice, and that's worth watching. In the meantime, if you're running a company, assume regulation is coming from somewhere, Europe, a big state, a judge, and get ahead of it: be mindful with employee and customer data, skip the surveillance-style tools, and get real permission before you use anyone's information.
Listen today on Spotify, iTunes, YouTube, or wherever you get your podcasts!
0:13 Why nobody trusts AI companies right now
0:36 Why federal regulation is stalled for the next two years
1:50 Big tech's lobbying grip on Congress
3:04 The administration's pro-corporate tilt on privacy and IP
3:26 How the EU is leading with disclosure and watermarking rules
4:34 State and local bills, from California to the data center fights
4:55 Civil suits: how individuals are reining in Meta and Anthropic
5:44 What corporations should do to get ahead of regulation
Episode 11: Seema Alexander on the Great Business Rewiring with AI
A filmmaker showed her AI could turn six months of work into minutes.
Geoff Livingston August 27, 2026
Seema Alexander is the founder and CEO of Disruptive AI, the host of the AI CEO podcast, and the alumni chair of DC Startup Week. Four years ago, right as ChatGPT hit, she was asked to advise Magic ML, one of the earliest agentic development platforms, later spotlighted at Google I/O.
As part of her efforts, Seema worked with a cinematographer trying to compress six months of storyboard work into minutes, chaining ChatGPT, Midjourney, and Runway ML together. She told me it was the moment she knew AI was "10X bigger than the advent of the internet."
Seema has not looked back since, and now calls this stretch of business history the great business rewiring. She notes how implementing AI is surfacing dysfunction: the corporate silos, the legacy systems, the initiatives that never quite got off the ground.
Watch or listen to Seema on YouTube! Don't forget to subscribe!
Today, Seema is building intelligent operating systems that connect the unstructured data companies already have, including a current project supporting portfolio intelligence for one of the largest international development banks in the world. She is also building a portfolio of AI native companies: One using AI built on human synthetic data to disrupt qualitative market research on the way to giving humanoid robots emotional intelligence, another building adaptive, personalized commerce through a company called Bridge.
Her hope for AI is grounded in her kids, ages twelve and sixteen. She is betting the majority of people will use this technology for good, that it will finally move the needle on precision medicine, and that it can strip enough friction out of work that a single person and an AI agent can build a real business.
Seema’s new conference, AI Business Con, runs March 2 and 3 at Fuse in Arlington,.
Listen today on Spotify, iTunes, YouTube, or wherever you get your podcasts!
0:45 Meet Seema Alexander, founder of Disruptive AI and host of the AI CEO podcast
1:26 The moment AI clicked: advising Magic ML and watching a cinematographer compress six months into minutes
4:37 Why she calls this moment the great business rewiring
5:39 Building intelligent operating systems, including a portfolio intelligence project with a major development bank
7:15 Her portfolio of AI native companies: human synthetic data and adaptive commerce with Bridge
9:42 Her hope for AI: precision medicine, less friction, and businesses of one
12:49 AI Business Con and keeping humanity human as AI gets smarter
13:29 Where to find Seema and register for AI Business Con