Marcus Rhodes

Human flourishing through commerce and technology. Optimistic about the possibilities of the decentralized future.
DMs, collaborations welcome.

[Marcus Rhodes @MarcusR_67]

I consider myself rational and skeptical, and not utopian, but I am fundamentally an optimist. Technology and free exchange aren’t perfect, but they’ve massively improved our ability to live, and there is every reason to believe that this is continuing every day.

Optimism is a choice, but in the modern world it’s a superpower.

I expect to engage on AI, cryptocurrency and blockchain, fitness and healthy aging, Socratic and Stoic philosophy, history, and meditation.

I value skepticism and critical thinking over consensus. My professional work involves critical appraisal of scientific data, and exploring uncertainty using Bayesian methods, and this will undoubtedly come across in what I share. More generally, we can all be wrong, but being humble enough to recognize it is how we learn. There is great power, and great possibility, in allowing myself to reconsider my opinions.

I am minimizing the noise around the things I’m interested in. Don’t be offended if I mute or ignore political tribalism (Republican, Democrat, or anyone else), most current events, rage bait, complaining, despair, fantasies about the apocalypse, conspiracy theories, and pointless arguments.

[Marcus Rhodes @MarcusR_67]

This is fair. Biological aging also plays a role. It's understandable to hustle and grind a lot in your 20s and 30s to establish yourself, and it's easy to let health and fitness slide while you're doing that. But once you turn 50, there can be a heavy price to pay for eating like crap, not exercising, too much stress, bad sleep, etc. It's also a good time to start thinking about life after your career, something a lot of retired people are not psychologically ready for.

[Marcus Rhodes @MarcusR_67]

Maybe, just maybe it's ok for a public figure that most of us don't know personally to be opinionated and (to some tastes) abrasive. Other past and present people whose accomplishments are undeniable, even if I probably wouldn't want to share an apartment with them: Beethoven, Hemingway, Steve Jobs, Edison, Tesla, Andrew Carnegie, Churchill, Patton, Bob Dylan, Miles Davis, Florence Nightingale, Frank Lloyd Wright, Martin Luther, and many, many scientists.

[Marcus Rhodes @MarcusR_67]

The algorithms on social media really, really don't like balanced, nuanced views on anything, unfortunately.

I'm also very optimistic about AI, but I don't subscribe to some of the wild utopian fantasies about it that I come across online. I believe it will create enormous prosperity and opportunity, and I expect that (like every other major productive innovation) it will create more jobs than it destroys. But I do not expect, e.g., an end to all human illness, aging, or death. Nor do I think Elon's post-scarcity post-jobs techno-communist utopia is coming. Though of course I could be wrong.

[Marcus Rhodes @MarcusR_67]

I think about this story in the context of other well-known serial liars. Paolo Macchiarini comes to mind. I think there's something in this kind of person that compels them to keep escalating, to tell increasingly outrageous lies to see how far it can go. But they get away with it for 3 basic reasons:
1. If people really want to believe a story is true, whatever the reason, they can talk themselves into believing a lot of things.
2. Most people tell small lies, but can't imagine telling a really outrageous lie. They believe they'd be called out and exposed immediately for such a thing. Paradoxically, this makes us more likely to believe that a wild story is true; we can't imagine lying about something that sounds incredible.
3. It's easy to live in denial when a con artist is exposed, because no one wants to look stupid and admit they were taken in by something that looks like obvious fraud in retrospect. Especially if not looking stupid is central to your personal and professional identity, like most academics, bureaucrats and politicians.

[Marcus Rhodes @MarcusR_67]

They did so because these institutions don't exist to expose or curb abuses of power by the powerful. Consciously or not, those making careers in these institutions take it for granted that the powerful are a special class whose judgment is not to be questioned. They exist to confer legitimacy on whatever the powerful want to do. You won't get a better example than the non-reaction of self-described civil libertarians during COVID lockdown.

These people understand that their job is to appear to have integrity, not to actually have integrity.

[Marcus Rhodes @MarcusR_67]

@cremieuxrecueil True. It's easy to take our free speech here for granted as the shitshow of Twitter circa 2017-2020 recedes further back into the past. But we really shouldn't forget.

[Marcus Rhodes @MarcusR_67]

I suppose this is one of those "different strokes for different folks" things. I don't really get what you're still chasing if you're already that successful.

I had a relative (long deceased at this point) who was very, very wealthy; he was a corporate CEO/captain of industry type in New York in the 70s and 80s. Very successful in a boring industry (boring even in the 70s). But if I told you his name you'd learn almost nothing by searching online, because he was in no way a public figure. And I can confirm this much from his example: being wealthy and NOT famous is a fantastic life.

I would also refer you to this:
11 Reasons Not to Become Famous (or "A Few Lessons Learned Since 2007") - The Blog of Author Tim Ferriss https://t.co/VIEPvvOU53

[Marcus Rhodes @MarcusR_67]

I see a lot of comments comparing the addictiveness of AI to that of social media. AI can be addictive, but in a qualitatively different way: there's no equivalent experience to doomscrolling when using AI. Social platforms (including this one) encourage passive consumption of content, which habituates a passive, "external locus of control" mindset. With AI I feel like my curiosity is rewarded; it gets more interesting the more questions I ask. I don't feel mentally lazy when I use AI.

[Marcus Rhodes @MarcusR_67]

It's not just that he's not an elected (or appointed) official or an NIH scientist. He also has no medical training or any relevant research experience. This is literally just a billionaire buying access to a bureaucracy, and using it to bypass normal representative government to see his preferred policies enacted. The sort of thing Democrats like to pretend to object to.

[Marcus Rhodes @MarcusR_67]

Sort of. This is NOT a defense of Arday or Cambridge, but a couple things are worth pointing out:
1. The history of well-known serial liars and sociopaths suggests that these people are well aware that bigger, more outrageous lies are paradoxically more likely to be believed. That is, they're able to get away with lies like this because most of us implicitly assume that no one would tell lies so egregious.
2. It's easy for even smart, well-intentioned people to believe in lies if they really, really like the story. See Paolo Macchiarini and Elizabeth Holmes for other examples.

[Marcus Rhodes @MarcusR_67]

My general impression is that we'll get there. The mid-90s Internet was honestly pretty useless and still caught on.

But I definitely agree that more thought needs to be given to the consumer experience, and making it more practically useful outside of coding. I've had plenty of "holy crap this is amazing" experiences with AI, but Tesla FSD was the only one I didn't have to work hard to discover.

[Marcus Rhodes @MarcusR_67]

@audrlo For me the best current example of successful consumer-facing AI is Tesla FSD. You don't need a high level of technical knowledge to understand it or use it, and the value is very apparent the first time you try it.

[Marcus Rhodes @MarcusR_67]

Fun story: my cousin and her husband bought a huge, beautiful piece of land with a crumbling wreck of a house on it. Part of how they made the math work on a custom home: getting a tax break from local government in return for letting the fire department burn down the old home as a training exercise.

[Marcus Rhodes @MarcusR_67]

Also brought to us by the people who brought railroads, electricity, etc: eugenics (including forced sterilization), nuclear, chemical, and biological weapons, prohibition, colonialism, world wars, murderous totalitarian governments, Jim Crow laws, housing scarcity, and urban blight.

High modernism is OK so long as you believe the state should essentially be unconstrained by the rule of law to pursue whatever experiments it likes, and that its priorities should override pesky things like human rights, representative government, and personal autonomy.

[Marcus Rhodes @MarcusR_67]

I'm reminded of the sequence in "Silicon Valley" sometime in season 3 or 4, when Pied Piper finally builds exactly what they want to do, and the platform fails because customers don't understand it. I'd guess a lot of the public hear about agents and think "that sounds sort of cool, but I don't know how to set it up or how to use it". Outside of people with a ton of technical expertise, agents still feel like an interesting solution without a clearly defined problem to solve, and without an intuitive way to get started.

[Marcus Rhodes @MarcusR_67]

@cafreiman Does anyone else remember the minor panic over bank teller jobs when ATMs became widely adopted?

Narrator: ATMs did not lead to waves of unemployed tellers. They made it cost-effective for banks to open more branches, which created a need for even more tellers than before.

[Marcus Rhodes @MarcusR_67]

@theojaffee Partly it's SF being particularly in demand, but yeah: Oakland has a well-deserved reputation for high crime. When I worked in SF I had coworkers who lived in Oakland, and even those in nicer parts of the city (up in the hills) seemed to experience a ton of crime.

[Marcus Rhodes @MarcusR_67]

Honestly, I see the sense in both sides of the whole-body scan question. I certainly understand the skeptics: having more information *can* help you make better choices, but you have to know how to use that information effectively. And many people (including MDs) struggle to wrap their minds around concepts like lead-time bias, probability, and false positives. So it is likely that the tech would create a ton of anxiety in patients, a lot more work for MDs, and a lot of ultimately unnecessary care. The technology CAN be useful, but this will require research on whole-body detection that doesn't exist yet.

That said, there is something to the accusation that docs dismiss the technology too quickly because its use isn't well understood yet.

[Marcus Rhodes @MarcusR_67]

Interesting. I've always been skeptical of Blue Zone/health habits research, for a different reason: it doesn't follow that the very common habits of a tiny minority of really healthy people are responsible for their good health, because you aren't observing the (possibly large number of) people who had all the right habits but didn't live as long. And you need those data to know if the centenarians really are doing anything special. But I hadn't thought about this issue before.

[Marcus Rhodes @MarcusR_67]

Semi-serious general AI question, prompted by your medical treatment example: considering @elonmusk's point about optimizing AI for truth, is there a risk of making an AI go insane by asking it to make choices within the frameworks of dysfunctional bureaucracies that make no sense? Asking AI to make recommendations about medical care - outside of pure private-pay areas like cosmetic or LASIK - in the context of modern healthcare bureaucracies feels a bit like asking HAL9000 to take the Discovery One astronauts to the monolith without revealing the mission.