Tag: Artificial Intelligence

Categories
Internet Culture Startups

Day 1805 and Dark Leisure, Time Violence & Outputting Value

Any other software developers out there remember the mythical man hour? It comes from Fred Brooks’ classic book The Mythical Man–Month which argues that adding more people to a late software project often makes it even later. This is also known as Brooks’s Law.

The man‑hour is “mythical” when tasks are not perfectly partitionable and require significant communication, shared context, and integration.

I think in the age of artificial intelligence we need to be revisiting this classic complexity insight as it applies across a world where we understand even less about how the time of input drives its notional value.

Measuring productivity in hours is a relic of a past labor era. And most workers have little incentive to improve output when they aren’t paid for it.

If we had quiet quitting during the pandemic where jobs could be done in minimal ways without getting fired, in this new artificial intelligence roll out we see another type of value capture mismatch between input labor and firm.

We are seeing what Fabian Steltzer calls Dark leisure. Others call it shadow user innovation.

Innovation happening through employee adoption of new technologies that is opaque to management doesn’t get counted and workers are reticent to be transparent.

the reason ppl hide their AI use isn’t that they’re being shamed, it’s that the time-based labor compensation model does not provide economic incentives to pass on productivity gains to the wider org

so productivity gains instead get transformed to “dark leisure”

Fabian Steltzer

Anthropic released a study on the supposed stigma attached to using artificial intelligence at work. Humans are already reacting to artificial intelligence as if it were an existential threat.

Except it’s been generally existentially freeing up to this point. Anyone who has used commercial large language models on healthcare can attest to that. So why are hiding its use?

Even coders are doing it. And who can blame them. It’s a lot less fun for some folks to coordinate a swarm of agents than it is to write code for a living. If you wanted to be a product manager, well you’d already be one.

The boss makes a dollar and I make a dime so that’s why I prompt on the company dime!

We are seeing the early artificial intelligence era take off collide with industrial-era systems of management that are no longer relevant in age of increasing complexity.

We’re putting intelligence into systems designed to measure hours and surprised when there is a misalignment. A Twitter mutual has a theory of consciousness systems they believe makes this is a form of time violence.

Human beings can tolerate NP hard moments of complexity, but cannot survive continuous low-grade complexity

The gap between human adaptability and systemic inertia is now wide enough to generate an entirely new form of harm: time violence

– Idea Nexus Ventures

We just cannot keep up with the varieties and types of complexities that are arising, so any advantage that can be used is being used. And you’d want to hide that advantage as long as you can. Sharing it has no rational basis. I find that disappointing.

I’d rather we not vice signal artificial intelligence as it only harms us. The value capture won’t always match up, but the gains to be made are worth having so keep using it where it works for you.

Categories
Biohacking

Day 1804 and 5 Years of Whoop

Tis the season of “unwrapped” data round ups as technology companies owe remind us that they have very capable data teams without unnecessarily freaking out users.

The marketers get all the glory, as everyone figures it’s a few MySQL queries gussied up with a creative hook but I always wonder what happened to the great data and behavioral science fad of the early teens. We’ve moved beyond IFTT and Zapier into natural language vibe coding. It’s a time of incredible change in understanding the data we’ve lived inside for years.

I got my fifth year of Whoop end of year marketing when I woke up today. Which was a little annoying as I wanted to see my sleep from last night but who wants to rush through the end of year review right? Finding it again can be tricky.

I keep wondering what more can we get from the hardware that knows us best. I’ve given half a decade to Whoop and mostly what I’ve learned is Covid really took 20 points of my heart rate variability.

I’ve been frustrated of Whoop’s progress in the integration of a chat AI as it seemed like any commercial LLM could analyze screen shots of the Whoop as well as their own chat did with the data it already. But tenure counts for a lot and it’s still the tool we have all relied on for half a decade now. Or at least 2% of us have worn it everyday for that long.

Clearly the data l could be used in many more way more than the management team ever seems to enable their data teams to do regularly. And I’d like to get more out of my Whoop. So being reminded of the length of use makes me wish they would deliver on that future before another year goes by.

Categories
Aesthetics Media

Day 1802 and Very Julian Fellowes Coded

The temptations to build an investing case around a historical parallel cannot be avoided. Americans love their booms and busts. And we love grand television dramas about them.

Julian Fellowes is the stage name of a conservative British peer, actor and dude who gets BAFTA award for making television about aristocratic families familiar to adapt and Americans bailing them out.

Then he went on to make a period drama about righteous industrialists in America called the Gilded Age which isn’t as iconic as as it’s not as personal since obviously a British peer won’t understand American mores.

I keep reading editorials about what the artificial intelligence boom most resembles. This week it’s railway booms and busts and the fortunes made and most. We’ve got dueling mandates for skepticism and boosterism.

It’s just a little weird to think that we’ve already made the Silicon Valley drama about the last boom and bust moment and it didn’t get written by a British conservative peer but by a Gen Xer Mike Judge.

Maybe in another generation on Netflix we will get a sweeping historical drama about a polycule group house in San Francisco as the next Downton Abbey.

Categories
Culture Politics Startups

Day 1796 and I’ve Got Billions in My Inbox Julie!

I’m not new to the boom and bust cycles that have defined not only technology startups, but American herself. Most millennials have opinions about their malign status in an economy designed to borrow from the future for a dubious present.

Much of the world is in a state of panic over “the churn” of the old rules changing and the new ones not being quite clear. But it’s really not clear what happens next.

I think anything goes as the networks speed up our connections to each other through artificial intelligence. The end of the age of scaling means it’s time for the era of deployment is it not? Or are none of us Carlotta Perez fans.

I enjoy speculating as is the fashion. Do I think corporate debt financing of data centers is some time bomb in private credit? Not really, no. I think it’s way more likely that don’t understand the full demand case for coordination in a mediated world.

I don’t know if we can meet the demand to be perfectly honest. I will say I am way more worried about us not meeting the moment. Changes to our cultural environment are as hard as our material ones.

If I had to read sentiment, I’d say that everyone is absolutely sick of having their attention used like a fiat currency. We cannot inflate our capacity for focus as easily as we can inflate the dollar. And we will demand simplicity by any means necessary just to exist. And artificial intelligence will smooth our world to manage with what we’ve got.

I think running a decentralized world will prove to be far too complex for most humans and it will be mitigated by layers of choices in governance that will probably not always maximize for the freedoms we’v come to expect from the liberal world order.

And yeah I think we will need a lot of data centers for that coordination effort. That the state might be the ones with the most demand seems a little rich though. Every individual on earth will want to be on the right side of the ratings. That’s more network state than state and it will be a longtime horizon.

I know it doesn’t sound great on its face. And yet I think it has had upsides. The demand for real businesses that operate in some world of efficiency has never been higher.

And to some extent, I believe that was always the entire point of computing. Make things so much better and cheaper we move on to bigger projects.

Giving you video games and porn might have been a weird way to get to Mars but medicine is as driven by vanity as much as survival so I don’t judge reality. I just want us to get more nuclear power. I don’t ask for much.

We didn’t want a legion of information processing professionals. We wanted to change the material conditions just as the Industrial Revolution did. The invisible hand is a strange thing.

I expect we will see quite a bit of opposition to the people believe that we need more energy, more industry, and more science. The future and its enemies are legions. I always did find it funny that fashion critics had a better read on the future than anyone else. Virginia Postrel and William Gibson both have good taste.

Categories
Community Startups

Day 1792 and Grateful for The Exceptions

This Thanksgiving I am feeling particularly grateful for the exceptions in my life. My world is filled with the exceptionally rare. Rare people, insights, businesses, and outcomes are part of building something genuinely new.

I suspect I’ll have to justify my faith in investing in and introducing new technologies to the world. We are doing a lot of looking back as the path forward looks so uncertain. And I continue to advocate for looking forward with optimism.

We have a lot to integrate and metabolize into human cultural life. We will be forced to address these changes as they change our institutions and expectations over the next few decades.

There is a lot to dislike about the technology industry at the moment. We’ve evolved far beyond “startups” being scrappy zero to one experiments in the proverbial garage. Startups turned into “Big Tech” and that concentration of influence and money has not always lived up to the high expectations we have for power.

We have had multiple cohorts of businesses as a mature industry. And indeed we’ve had multiple generations of people who spent their entire lives building a global ecosystem of technologies, along with the talent and capital to scale it. We may relentlessly start afresh but we cannot avoid acknowledging that we are a power base in our own right now.

Just in my lifetime, we’ve publicly codified our cultural mores, shared decades of knowledge on best practices on the open web and built institutions dedicated to helping people work across the multiple fields and disciplines that encompass “technology” as an industry. Or maybe I should simply call it an economy. It may even be the economy at this point.

Which is a problem. Our capital sorting mechanisms have seen our efficiencies and returns and pushed more resources, human and financial, towards us.

That has frustrated and starved the industrial base that provides us with the infrastructure to build. Let’s not even get started on what it has meant for food, education, entertainment and family.

I began more seriously investing in startups at the beginning of the pandemic. We maintain a small fund with low key LPs and our own family capital.

That is enabled by what we jokingly call the circle of life that is a liquidity event. When a startup sells many people become not just a little bit better off but sometimes twenty or even hundred times better off.

Those outlier events pay for all of the other things which don’t work as well. It’s a hits driven business. Hollywood would say “Thats show biz baby!” Oddly we don’t have a simple way of explaining the randomness of who or what becomes a winner.

Being excellent just isn’t enough. Startups that succeed are often exceptional in all areas and even then it still might not work. That bothers losers more than it does winners because the winners can comfort themselves with the money. But deep down even the winners know it could have easily gone another way.

So this Thanksgiving I am grateful for all the exceptional cases that have come into my life. To even see one is a rare thing. To be exposed to dozens of them is extremely unusual. To be invested in even positive outcome from the very start is beyond rare.

We’ve done so much to make startups more accessible to those with the mindset and discipline to succeed and still so many barriers remain. I see my work as the first check a founder takes as being a small part of the cycle of exceptionalism that builds success.

Just in the past two weeks we’ve had three companies raise large scaling rounds at markups that now place them soundly in the exceptional category. In two cases, I was their very first check, and in the third I was in their first pre-seed round. I qualify it only because I was not the first person to commit which I strive to be.

That is where I strive to be exceptional. I want to be the very first person that sees you for what you will be.

And I am deeply grateful to the founders that allowed me to be their first believer. It’s hard to be a founder. I’ve done it. To be an investor is much easier. You just have to have the balls, the brain and the bravery to say “yes” to something nearly impossible. That I can say yes is something for which I am most thankful.

Categories
Internet Culture Media Politics

Day 1785 and Adversarial Openness

There has been quite a bit of discussion in alignment theory with artificial intelligence that considers how legibility and openness might work at cross purposes when coordinating across different intelligences with different goals. Politics exist everywhere it would seem.

If you are transparent, but seek to change an agent’s behavior, you might reasonably be interpreted as adversarial by the agent. So it follows you must consider that their actions are no longer collaborative and open towards you but potentially adversarial and opaque depending on how it judges you.

The information habits and “winner’s optimism” that some American elder millennial display in public digital spaces are telling. In particular, we have skewed heavily towards legible openness as our internet was often friendly and our geopolitical positions was dominant.

These conditions are no longer true. And so we are now experiencing the Dark Forest Theory of Yancy Strickland (based on Liu Cixin’s stellar science fiction series the three body problem). American millenials are on a very different internet than we grew up on.

I’ll admit I have a bone to pick with Yancy as it felt more like he was defecting from the open web in 2019 because it was scary and filled with fascists. I didn’t think he believed it was because it was actually dangerous. His return makes me question his original declared intentions and his goals now.

The Dark Forest disappearing man has come back to the open web now. Things have changed and we all need our own private Idaho. Which you can find through his offerings.

I’ll note he needs the distribution channels of large adversarial networks like Twitter and that means gaining power in the dark forest. As we consider how open and legible to be in this very difficult moment I thought this was an instructional revealed preference.

Categories
Politics Startups

Day 1784 and Strange Trade Offs

It’s been a pretty fantastic few weeks for my investments. Decisions made years ago are now looking pretty smart. A bet I made two years ago announced a round and then proceeded to announce splitting the atom the next week.

Not to only focus on current belle of the ball in Valar especially as everyone in the portfolio seems to be finding their way. We are lucky that we focused on compute, energy, and decentralization as that is the trifecta of the artificial intelligence wave.

I honestly didn’t expect that we’d see such progress in our nuclear pick. With the regulatory climate it seemed more likely compute marketplaces and inference products would outpace the most regulated technology in the world.

Somehow during a Trump administration you get unexpected outcomes. I’ve been fighting for compute figuring the energy bottleneck wouldn’t get addressed till we had the full supply side of new AI products. It turns out everyone wanted to rush into capital expenditures and infrastructure as the demand was already there.

I guess I’ve proved my own thesis again. You can get a read on the direction and maybe even first order effects but in a chaotic world the second and third order effects are much harder to predict.

And on balance for all the bad I think on balance the atomic age finally arriving might be a worth while trade for our future. Hard to say if I’ll keep that opinion but I am grateful America is getting back on track with nuclear power.

Categories
Community Internet Culture

Day 1780 and Being Cooked

I’m starting to think the more optimistic you are about the future, the more cooked you think we are. I didn’t expect this.

The Doomers have a coherent worldview. It’s simple to imagine involving losing your humanity to machines. This is at least legible and a call to our common humanity. Change is scary and bad and we don’t know how any of this is going to go. So why not be cautious?

The optimists are all excited about different things though. And that opens us to a lot of attack paths. And yes I’m calling myself an optimist though I have a lot of downside scenarios on my radar.

Some of the outcomes that you might find dystopian are the utopian outcomes for someone else. Think Caliphates or Communist surveillance states.

The complexity of our reality is so far beyond the grasp of your average person it seems cruel. And we sympathize with the struggle to adapt because it appeals to our common humanity.

It’s no wonder America has had so many revivalist movements. We have changed so much in our 250 year history, we are always rediscovering the value of faith. What else do you have when the future is uncertain? If we are cooked anyways we may as well all take Pascal’s Wager?

Categories
Politics

Day 1763 and Baumol’s Cost Disease Accelerationism

Today was a pretty big news day. It was a FOMC meeting with a cut, Jerome Powell gave some forward guidance that a cut in December is not guaranteed (cue market upset), and NVIDIA became worth $5 trillion.

This is apparently 16% of our GDP and without investment in artificial intelligence related build-out, our economy would have only grown by 0.6%.

Without Magnificent Seven spending, GDP would have grown at a mere 0.6% annualized rate instead of around 1.1%-1.2% – Fortune

So America would be looking about as gnarly as Europe without the Magnificent Seven and AI infrastructure build-out spending.

About 92% of GDP growth in the first half of 2025 was driven by investment in data centers, AI infrastructure, and information processing, with NVIDIA as a primary contributor Yahoo Finance

Which is a scary large amount for any corporation, but is somewhat rational in the logic of a civilizational technology changeover akin to the Industrial Revolution.

For some comparisons, Standard Oil at its height represented about 5-6% of the total U.S. stock market value at the time and 1.5% of America’s total GDP. AT&T’s Bell Systems were worth about 3-4% of America’s GDP at their asset peak in 1984 so not entirely an unprecedented situation though Nvidea’s percentage is a very networked era problem.

How afraid should we be about the potential for a market bubble in artificial intelligence? That is a questions for Carlotta Perez

Having lived through both the dot-com crash and the global financial crisis, I have some fears, but also this feels about as rational as any of the other ways we’ve handled valuations and value in past boom-and-bust cycles.

There is significant revenue from very real demand. It is just hard to see the demand as it’s industry demand not consumer. And the consumer demand we have is likely coming from professionals who are more enabled in ways we can’t count. I couldn’t have answered half the questions I had for this post before the LLM age.

And that demand for efficiency was coming and needed to be addressed over some time horizon, no matter what.

As different industries cope with their extreme lack of efficiency in the face of other industries who are efficient and in demand wages rise everywhere and basic needs like education & healthcare get more expensive despite not being delivered more efficiently.

So we still need those inefficient industries but what do we do? We have to find solutions.

Because we were going to need to build out the infrastructure for diversified energy transition. Much of this is being spent on build-outs for things that we genuinely need.

We need nuclear. We need power grids that aren’t from the dark ages. We need the efficiency for compute as government services have gone full runaway Baumol accelerationist. Unless we do the hard work that’s going to take 10 to 15 years, most liberal economies will collapse under the weight of the social safety net.

So we need to do a fairly thorough job of investing in the future, independent of whether it’s artificial intelligence driving our future or developing an industrial policy of, say, going to war with China. Necessity is the mother of invention and I’d rather the need be capital growth than war to drive industry.

I don’t know why this “facts of budgeting life” works people up so much. Booms and busts and bubbles build real things and we really need more efficient energy, healthcare, and education.

The economy is a nutrient gradient and money moves to where it gets fed. Right now the promised efficiency of a solution to unsustainable spending is paid for by gains in areas which did get more efficient. That is just the whole game. Grow faster and bring along anything that isn’t for the ride.

Categories
Biohacking Medical Startups

Day 1753 and Vibe Coding for Your Skin Health

Many moons ago, when I was first attempting to get a diagnosis for why I was always in pain and exhausted, I got a battery of allergy tests. I did the “gold standard of allergy testing” called patch testing which is a form of pin prick testing designed to pick up responses that may be delayed.

It was an awful experience. I barely made it through the 5 day trial between the 100 allergen pin pricks and final measurements.

I remember begging the doctor for a way to measure early. I asked if I could take some Benadryl to take the edge off. Alas the only way it would be accurate and covered by my insurance is if I gutted it out.

You are not allowed to shower, sweat, be exposed to UV rays (no going outside) or take immune suppressants that might subdue your body’s response.

I was struggling to breathe, my entire body itched and ached, and I had a migraine so bad I couldn’t see for the stars & dizziness. It’s possible I wasn’t stable enough to have adequately consented to the test but I did get my final results.

Out of 100 common allergens tested it was confirmed I was extremely allergic to 10 of them with another moderate sensitivity set of twenty or so that I should merely try to avoid as opposed to my firm “no go” list.

The dermatologist gave me a sheet with 75 different chemical names and formats that I might encounter in the wild from these core allergens:

Budesonide
Ammonium Persulfate
Benzisothiazolinone
Limonene
Oleamidopropyl Dimethylamine
Formaldehyde
Lauryl Glucoside
Methylisothiazolinone
Propolis
Thimerosal

I instructed to search ingredient lists for these names any time I purchased a household product, personal care item, cosmetic or other item which might include these ingredients which ranged from nail care to vaccines.

It was honestly quite overwhelming. And some of the above ingredients are in basically everything. I dare you to avoid Limonene for a month.

So my husband and one of our best friends did what any practical minded engineer would do and they made me simple Google sheet where all 75 varietals could be checked if I plugged in the ingredients from any item.

I used it for years. I’d plug in the INCI from every brand I encountered into the sheet no matter what. I gave away a lot of products to friends.

Today it occurred to Alex that we should probably vibe code the thing into a proper web application using Replit so other people could check ingredient lists for their own allergies.

Within two hours, most of which was waiting on the kindly AIs to do their thing, we had a fully functional web application.

You can set your own allergens or click a few buttons for common allergens and “clean ingredient standards” and run a check for an all clear.

It isn’t super fancy but it doesn’t need to be. It just needed to keep your data safe, be easy to log into so you can securely check and access your personal list and generally functional enough to change and set allergens. We’ve put it on our own little domain just to see how much this will cost to run (and we’ve set up alerts so it doesn’t go bonkers) but we figured this should be accessible and simple.

And while there are other options on the market, most are bloated, overly paranoid and designed for scaring California moms rather than quickly helping people with clear preferences for avoidance and actual tested allergies. So hopefully our pain can help you breathe easier.

Some options for chemicals and irritants you can select on our app.
My own settings of allergies and sensitivities