✦Crazy Wisdom

Stewart Squared

Episode #52: The Illusion of Choice in Big Tech

Episode #52 · August 28, 2025 · 52 min

MP3 · Apple Podcasts · Spotify

About this episode

In this episode of Stewart Squared, Stewart Alsop sits down with his father, Stewart Alsop II, for a wide-ranging conversation on the frustrations of modern UI/UX, Microsoft’s struggles with spam and AI adoption, Google’s approach to knowledge management, and the broader lessons of technological hype cycles from fiber optics to GPT-5. Together they explore how big companies evolve from serving programmers to serving enterprises, touch on the role of regulatory capture in shaping user experiences, and recall stories of early email, Hotmail, AOL, and long-distance calls in the 1960s. Along the way, they connect today’s debates on monopolies, Bitcoin, and satellite internet with personal anecdotes from their family history and reporting trips to Moscow.

Check out this GPT we trained on the conversation

Timestamps

00:00 UI/UX frustration, Microsoft spam vs Gmail; scam email triggers rant on filtering and usability.
05:00 Admin controls, external IT friction; Google Drive knowledge management and closed-by-default files.
10:00 Bitter lesson, compute at scale; GPT-5 hype, model consolidation, tokens and cost signals.
15:00 Consumer UI simplicity vs programmer leverage; Bitcoin early-adopter edge; Coinbase code alerts.
20:00 Regulatory capture thesis—Microsoft, Coinbase, Palantir; too big to fail, users sidelined, startup opening.
25:00 Monopoly talk: Netflix, Apple App Store; success metrics and venture-scale outcomes.
30:00 Microsoft arc: programmers → enterprise; MS Basic, MS-DOS/Seattle DOS, IBM; latency woes on the call.
35:00 Starlink Mini portability, power limits; satellite iPhone messaging; T-Mobile, Globalstar arrangements.
40:00 Email history: AOL, CompuServe, Hotmail/Yahoo; Gmail scale; Outlook/Office 365 vs Edge/Safari.
45:00 NEA standardizing on Windows, regrets; Riverside recording hiccups; early Gmail usernames, scale effects.
50:00 1963 operator calls, injury story; Moscow reporting trips; Khrushchev–Nixon Kitchen Debate context.

Key Insights

  1. Stewart Alsop and Stewart Alsop II opened with frustrations around UI/UX and how even industry leaders like Microsoft fail to implement effective AI for basic tasks like spam filtering. Gmail adapts instantly to user feedback, while Microsoft’s Exchange requires convoluted admin settings, leaving everyday users powerless.
  2. Their discussion shifted to Google’s knowledge management problems, highlighting how file access defaults in Google Drive create needless barriers. Both observed that corporate bureaucracy shapes user experience more than technology itself, reflecting how large firms prioritize control over usability.
  3. The “bitter lesson” by Richard Sutton framed the conversation on AI. The Stewarts compared today’s trillion-dollar GPU investments to the fiber optic overbuilding of the 1990s—misguided methods that still laid crucial foundations. They questioned whether GPT-5’s consolidation into one model was a sign of efficiency or hype masking economic strain.
  4. A key theme was programmer leverage. They noted that programmers who mastered Bitcoin early became “post-economic,” while non-programmers remained locked out. This reinforced their point that tech often empowers a small, technically literate class while excluding ordinary users.
  5. They critiqued regulatory capture, suggesting Microsoft, Coinbase, and Palantir thrive not by delighting users but by embedding themselves with governments. Once companies become too big to fail, their true customers shift from individuals to institutions, and user needs fade from priority.
  6. The episode revisited Microsoft’s history, from buying Seattle DOS to serving programmers and then enterprises. They argued that companies inevitably drift away from their original users, though some, like Microsoft through its OpenAI partnership, manage to stay relevant despite this drift.
  7. Finally, they wove in communications history—from AOL and Hotmail to Gmail’s dominance, and even back to 1960s operator calls when family news was relayed across continents with constant dropped connections. These stories framed the present as just one phase in a longer evolution of technology mediating human connection.
Episode transcript
Stewart Alsop III00:06

Everyone is wrong about AI, but if you know the history of technology, you can be wrong and still be right.

Stewart Alsop II00:13

I know the history of technology since I started with the personal computer in 1980. So listen in while we explore what is right about AI.

Stewart Alsop III00:28

Oh, Riverside decided not to give us the five seconds intro. So here we are, another episode of Stuart Squared and we're going to be talking about some interesting stuff related to the fact that the current Internet is based off of an old model where humans have to navigate UIs and UXs. And now I'm not. Now the more that I know how to vibe code and the more that I know how to program UIs and understand all these little details of it, I get so infuriated. And I think that this is something I get from you. I am understanding that I complain about the software that I experience with on a daily basis and I think you do as well. And so we were talking about how Microsoft is this leader in AI, but

Stewart Alsop II01:21

they can't put it on can I share screen and have it be recorded? Can I share that screen I just showed. Yeah, so let's try that. So, yeah. So for our viewers and listeners. Listeners obviously can't see this. I'm sharing a screenshot of an email that I got today. One in five or six different ones, which is obviously a scam. They want you to click on a button so they can collect your data. And the thing, I mean, it's just badly titled, but it comes from an email that goes on forever with all these random numbers in it. And it's like obviously spam. And this comes at our to my allsuploui.com email address which has been widely distributed on all kinds of lists and stuff. So it's not hard to get it. And it says attention S alsop. It's like they don't even know my name. And this is Microsoft 365 and they have all this stuff, but it doesn't look like an email from Microsoft. So my question is when I get this is why can't Microsoft use ChatGPT and all the open AI technology, not to mention its own technology, to filter this out as a human being, I'm looking at it and I'm going like, well, this is obviously spam. So I mark it as junk. Does the junk marking go anywhere? Do they collect that data? Do they know now on Gmail, which is also. My email is pretty widely distributed. When I mark it as spam, it doesn't show up again. They instantly know that if I'm marketing as spam. It's something. It should be filtered out for everybody. And so I don't get very much junk in my Gmail. I've heard there's a difference on workplace accounts, but I'm an individual user but. So Microsoft has used its technology on behalf of its customers in a very unusual way because generally speaking Google doesn't do a very good job of that. But in Gmail they do. And Microsoft is clueless. I mean how can you believe that this company is the future of AI when they can't even filter out obvious spam? And there's a lot of other use cases. I mean it's this one that just infuriate.

Your first full transcript is free. After that, an email opens every transcript in the index — a list of readers we can write to, not a guest book.

Subscribe