Casey Sapp

Serial Founder; Investor; Agentic Engineering

[Dwarkesh Patel @dwarkesh_sp]

Human beings whose emotional centres are damaged, even if their intelligence is still intact, have terrible decision-making skills.

Whatever role emotions are playing in humans, it's necessary for agency.

Ilya speculates that the equivalent for AIs is something to do with value functions - and that it might not emerge through pre-training alone.

[Casey Sapp @CaseySapp]

Brilliant

[tuōmo @7uomoki]

Hitler running a frontier AI lab in SF hears about the "Car Wash Prompt"

[Thariq @trq212]

https://t.co/R6exTuF7P8

[Casey Sapp @CaseySapp]

A contact from Defense Innovation Unit (DIU) reached out and asked I️ disseminate this to my network and remind any defense focused AUV/ROV/UUV companies to submit a Capabilities Matrix for one of the 3 tracks in this solicitation by June 10th.

Link to Apply: https://t.co/KWi2t1uBLz

[Casey Sapp @CaseySapp]

One thing I’m noticing in the AI discourse as of late and that @bgurley is keenly making the case for on the @allinpod_clips this week is that @Anthropic may be following the Nick Land playbook more closely than most realize with his “Dr. Frankenstein” theory.

For those not familiar, Nick Land is a philosopher revered by some in Silicon Valley since the 90’s who has argued that technological forces will accelerate beyond human control and eventually direct the future of human civilization. Gurley assesses, and I think accurately so, Anthropic is pursuing a structurally similar outcome. The catch is that Anthropic’s vision is framed as benevolent, but structurally it is the same argument.

Anthropic presents it as a moral responsibility — build superintelligence safely and steward it responsibly — but unlike Land isn’t explicitly celebrating the eventual loss of human control.

Whether this characterization is fair or not, it strikes me as one of the more interesting lenses for understanding the emerging philosophical divide inside AI. Should we accelerate or restrain AI, and is it an autonomous systemic force or will it remain something humans actively steward.

High quality episode this week driven by the release of the Pope’s letter “Magnifica Humanitas”. I️ want to see more conversations like this.

And, Bill, Go Gators.

[Casey Sapp @CaseySapp]

One of the better takes I’ve heard in a long time.

[The All-In Podcast @theallinpod]

Bill Gurley: Anthropic Thinks It’s Building God

@Jason: It is the ultimate level of narcissism and delusion of grandeur to think you can create God.

@bgurley:

“Anthropic is a mystery to me. I've never, ever seen a company that is both leading their field and the most negatively outspoken commenter on what they do.

And my initial theory was the regulatory capture theory. Quite frankly, I think they're very close to achieving that.

But then they just got so loud that I've literally, in the past 30 days, read everything I can about Anthropic, and I've come up with a new theory.

I call it the Dr. Frankenstein theory.

The more I dig, I've met people who, I dare say, think it's their responsibility, and they're excited about, building a species that's superior to humans.

Dario wrote this blog post called ‘Machines of Loving Grace.’ It was based on a poem.

The last stanza of the poem says, ‘I like to think of a cybernetic ecology where we are free of our labors, and joined back to nature, returned to our mammal brothers and sisters, and all watched over by machines of loving grace.’

Sounds like an overlord to me.

And then in Dario's post, he says, ‘It could be a capitalist economy of AI systems which then give out resources to humans based on some secondary economy of what the AI systems think makes sense to reward in humans…’

So I don't think they think they're writing software. I think they're midwifing a deity here.”

Jason:

“These are delusions of grandeur. Let's call it what it is.

They believe that they're so powerful, these individuals, that they can create God, and that by creating God, they are like this Prometheus kind of species.

It literally is the ultimate level of narcissism and delusion of grandeur to think you can create God.”

[Casey Sapp @CaseySapp]

This is insane and awesome.

I️ have a special appreciation for how hard this is as I was building towards this in an underwater environment before Blue Ring Imaging sold to VideoRay in 2023.

[Anduril Industries @anduriltech]

The difference is night and day.

The digital night vision of the EagleEye Family of Systems delivers an 84 degree field of view, stereo thermal fusion to expose hidden threats, and a 4K display for enhanced warfighter perception. Double the industry standard.

Pictured: EagleEye (left) and standard PVS-31 (right).

[Casey Sapp @CaseySapp]

@carolynrogers Crushed it. Great writing.

[Casey Sapp @CaseySapp]

3 Stories That Explain the Opportunity with AI:

#1 Photography vs. Painting

When photography was invented, painters feared extinction. Why master realism if a camera could do it instantly? But photography didn’t kill painting you could say it liberated it. Artists have increasingly moved toward emotion, symbolism, abstraction, and imagination. The camera has forced painters to ask: “What can humans uniquely do that machines cannot?”

#2 ATMs vs. Bank Tellers

People thought ATMs would eliminate bank tellers. Instead, banks became cheaper to operate, more branches opened, and tellers shifted into higher-value relational work: advising customers, solving problems, and building trust. The role evolved upward.

#3 Milton Friedman’s “Spoons”

While visiting a canal project, Milton Friedman noticed workers digging with shovels instead of tractors. Officials explained it was a jobs program. Friedman replied: “If it’s jobs you want, why not give them spoons?”

Friedman’s point is the same - human flourishing does not come from preserving inefficiency for its own sake. It comes from removing unnecessary drudgery so humans can pursue higher-order work. This is the real opportunity with AI. Not making humans obsolete, but making humans more human again.

I suspect the societies that thrive in the AI era will not be the ones that simply automate the fastest. They will be the ones wise enough to distinguish between what should be automated and what must still be lived.

So here’s the litmus test I am increasingly using when engaging with AI: Is the task capable of forming me, or consuming me?

[Casey Sapp @CaseySapp]

Using AI for Human Development: Outsourcing the Journey Instead of the Drudgery

After having spent weeks listening to podcasts, reading books, and studying the arguments coming out of Silicon Valley I keep feeling like much of the AI conversation is missing something fundamental. Many of the loudest voices are either framing the problem backwards or barely discussing the possibility that AI could actually be deeply humanizing. The media especially likes to frame the threat that machines will make humans unnecessary.

I increasingly think the larger societal risk is that humans will voluntarily surrender the parts of life that actually develop them, because automating white-collar work does not inherently diminish human dignity or the soul. In many cases, it actually can be restorative in its nature — with conditions.

Why? A shocking amount of modern knowledge work is really just “cognitive ditch-digging”.

I keep coming back to a distinction that feels increasingly important to discuss in the AI era and the questions we should be asking of which to serve as the axiomatic filter in how we use it:

- Is the task at hand required primarily for the result?
- Is the task at hand important because the doing itself matters?

Two examples to drive this home:

#1 If you need a drainage ditch dug you just want the ditch. If a backhoe can do it in five minutes instead of eight hours, you will happily use the machine. The labor itself was never the point.

#2 Instead of traveling to Europe, you don’t want someone to implant fake memories of Paris into your brain while you stay home. The experience is the point - the wandering, getting lost, struggling through unfamiliar streets, tasting food, having conversations, etc.

This distinction explains almost the entire AI debate for me. The danger is not automating drudgery, it is accidentally outsourcing formation.

Formatting spreadsheets, writing repetitive corporate emails, producing decks nobody reads…this is not evidence of intelligence as they were never deeply human to begin with. They are administrative burdens executed on computers because better tools did not yet exist. They aren’t created to make us feel more alive. That comes from mentorship, wisdom, conviction, courage, raising children, pursuing truth, or creating beauty. Those are not “ditch-digging” activities. Those are journeys. And journeys shape the soul precisely because they cannot be shortcut without losing something essential in the process.

If you use AI to summarize a 600-page regulatory document, you are removing drudgery. If you use AI to avoid wrestling deeply with hard ideas, avoid developing conviction, avoid learning how to think, avoid writing, avoid moral struggle, or avoid the pain required for wisdom, then you are essentially asking for the memories of Paris without ever taking the trip. You get the output, but lose the formation.

Human flourishing was never about maximizing friction everywhere, much of it today is pointless.

Continued ⬇️⬇️⬇️

[Casey Sapp @CaseySapp]

@pmarca Amen.

[Casey Sapp @CaseySapp]

This is interesting: https://t.co/ZwYf84HKTS

[Casey Sapp @CaseySapp]

I’ve always been a builder, an executor, and a creator. I get far more energy from actually doing the work than from philosophizing about it. But as this AI “supersonic tsunami” continues to accelerate, I’ve found myself increasingly uneasy with the messaging coming out of Silicon Valley and the current batch of thought leaders.

The Elon Musk, Chamath Palihapitiya, David Sacks, Peter Thiel, and Peter Diamandis’ of the world are undeniably brilliant. They get asked constantly about AI and its role in society. But the answers are incomplete as they almost always frame the future through the lens of efficiency, optimization, productivity, and economic output — an Industrial Revolution mindset where the highest good is reducing friction and maximizing capability. It can be intellectually easy—even frictionless—to focus on the similarities instead of differences as AI is uniquely designed to imitate us.

I want to hear more discussion about how to cultivate wisdom in a world where superintelligence and knowledge is free, or if AI is a tool for the expert and crutch for the learner how do we ensure our kids don’t shortcut the process and become men/women of substance? I️ am looking for fresh perspectives that practically embrace AI and its benefits, yet understand matters of the soul.

If you have podcasts, books, research, or links to this subject I️ would love to hear recs. Not doomerism or techno-optimism, but something new.

[Casey Sapp @CaseySapp]

@sama Better responses to theological/spiritual questions.

[Casey Sapp @CaseySapp]

@drtaubraun @NotabotSir Has this been addressed in your recent posts? I️ may have missed it.

[Casey Sapp @CaseySapp]

Who else is noticing that US companies building ocean-based AI infrastructure may unlock entirely new compute models while terrestrial and orbital infrastructure struggle with scaling bottlenecks?

One of the biggest opportunities I’m seeing may be AI sovereignty-as-a-service. As more data center projects get delayed or canceled — and as governments globally tighten AI and data regulations — jurisdictional flexibility itself could become a premium product, similar to how Switzerland became a financial hub or Bermuda became an insurance center.

There are also meaningful defense applications. Floating or relocatable compute platforms could be significantly harder to physically disrupt, seize, or target than fixed terrestrial facilities.

Shipping fleets, offshore energy operators, subsea cable companies, and oceanographic researchers all generate massive amounts of data that is expensive to transmit back to shore. Offshore compute enables real-time AI processing much closer to the source.

But the biggest near-term driver is probably speed-to-market. If offshore AI nodes can be deployed in months instead of years, companies gain a major advantage during the current AI compute shortage. That deployment speed — more than cooling alone — may be one of the most important infrastructure opportunities of this decade.

A marine tech category worth following more closely.

[Casey Sapp @CaseySapp]

@drtaubraun What does “turn on” mean to the normal person?

[Casey Sapp @CaseySapp]

I’m highly familiar with our program in San Diego related to this and it’s awesome. I️’m sad it’s winding down. Also no kamikazes in the Navy fleet.

[Eric Daugherty @EricLDaugh]

🚨 LMFAO! SecWar Pete Hegseth responds to reports Iran will deploy "kamikaze dolphins" against the US:

"I can't confirm or deny whether WE have kamikaze dolphins, but I can confirm they don't!"