My New Hugo Site

Mon, 01 Jan 0001 00:00:00 +0000

AGI is a feeling, AGI is a game.

AGI is a feeling. We know we have AGI when we look at an AI system and feel like it is intelligent.

AGI is a game. We know we have AGI when Sam Altman says we have AGI. If Sam Altman says we have AGI and we don’t feel like we have AGI, then he’s shot his load too early and there’s no impact from declaring victory on AGI.

Mon, 01 Jan 0001 00:00:00 +0000

2 years ago, Sam Altman may have had the most prescient thought in the history of AI thought leadership when he said “I expect AI to be capable of superhuman persuasion well before it is superhuman at general intelligence, which may lead to some very strange outcomes.” (https://futurism.com/sam-altman-ai-superhuman-persuasion)

I’m going to invent a random “Automation Ability” Score to prove a point. Let’s say prior to GPT-4 and LLM’s, our ability to automate score was 10,000. Pretty good, we were automating a ton of shit, and let’s say now, LLM’s have an ability to automate score of 1,000,000. That’s such a big leap, that I don’t think the average human can understand it. It will take years, maybe decades to understand exactly what an automation score of 1,000,000 even entails. But what if the “virtual co-worker” role we want LLM’s to take on requires an automation score of 1,000,000,000. These numbers are so wildly disproportional, the average human can’t wrap their head around them. We don’t actually know the scores, because we don’t know how to measure them. All we see are the gaps. The gaps between what we could do and what we can do, and the gap between what we can do and what we want to do. We see that there are gaps, clearly, but that’s all we can see.

Mon, 01 Jan 0001 00:00:00 +0000

I’ve wanted to write this article for a while now. Because I often see the saying “Devops is not a role” all over the place, and every time I see it I have a visceral reaction.

Ironically, I’m reading the book “Team Topologies” right now, which seems like the ultimate “Devops is not a role” book. But I want to unpack my thoughts for a second. I’ve often noticed that when I’m presented with a new idea, my impulse is to argue against it, play devil’s advocate. Partially this is for the same reason most people do. I’m afraid of change; don’t move my cheese, and this reaction is usually not helpful. Playing Devil’s advocate without a strong purpose in mind is an anti-pattern.

Mon, 01 Jan 0001 00:00:00 +0000

https://www.pbs.org/wgbh/pages/frontline/shows/dotcon/historical/bubbles.html
The Dust Bowl

Mon, 01 Jan 0001 00:00:00 +0000

Left vs Right

I happened upon this youtube video, called “Why Intellectuals are Mostly Left” with Roger Scruton (the title was meant in a deragotory way)

This interview wanders into the idea “The Nazi’s were actually left / socialist policies”, and fine. But really, when you dig into this interview.

“You keep learning and you move to the left, and then you give up your learning and move back right”

Mon, 01 Jan 0001 00:00:00 +0000

Nostalgia is shallow.

Discussions about Nostalgia on the internet are often pretty shallow. “Everyone is nostalgic for the time of their teen years”. The free press ran an article called “The past wasn’t inherently better. So why does it feel that way.”

It feels like the zeitgeist is there are roughly two sides to the argument. Either you’re a backwards looking sentimental-fool who waxes for a time that is inherently tainted by rose-colored glasses, or you’re a realist who looks at the data and understands that lifehas never been better.

Mon, 01 Jan 0001 00:00:00 +0000

Every piece of advice is based on optimizing towards or away from a given impulse.

The reason influencers focus on the grind, fitness, entrepeneurship and the like, is that for most of us it’s an unnatural impulse. I don’t want to speak for everyone, but my most natural impulse is towards laziness. Definetely towards not doing. But the grind isn’t “good”. Not in an unqualified way. For the first part, if what you’re doing is not working, you probably need to grind less, step back and assess your strategy.

Mon, 01 Jan 0001 00:00:00 +0000

YAML is structured, String interpolation is messy, problem is argo since YAML is structured.

If you don’t invest in observability, etc, microservices will hurt and not help. LIkely, you want to invest in Observability automation.

https://www.youtube.com/watch?v=CFLqAn8kEcs

Starting with a purpose is important. Otherwise you get in the trap of “If I just bult the perfect framework, everything would be amazing”

Mik’s self note / principle -> When you’re evaluating two pieces of software, or trying to figure out what you’re doing, MAKE SURE YOU enumerate what are your goals, what is the why? Not just, “We should figure out whether or not Terraform is the right solution for us. What problems does Terraform solve, what problems do you have for it, and what do you hope a new solution will give you?

Mon, 01 Jan 0001 00:00:00 +0000

Propaganda

I imagine that people who live in North Korea don’t all accept the official narrative.Certainly, no one is allowed to project opposition to the narrative, and that must be really awkward, being surrounded by people who won’t admit they don’t believe the official narrative, and may even turn you in if you admit you don’t believe the official narrative, but everyone is walking around thinking “this is ridiculous”.

But in America, for propaganda to be effective, people have to want to believe it. It has to support some narrative or emotion or idea they have about the world. It’s not that it has to be strictly speaking believable. Lots of outlandish things were said about Hillary Clinton during the 2016 election, and certainly not all those things were believable, but in those cases, it fits pre-existing beliefs about the world. “Maybe Hillary Clinton didn’t do thing A, but she’s the type of person to do thing A.” Same thing when the killer of Charlie Kirk’s politics turned out to be complicated. I’m guessing many folks don’t know those politics are complicated, nor care, because it’s directionally speaking “correct”. I “know” that the Democrats are violent, therefor if they didn’t commit this act, does it really matter? They’ll commit others.

Mon, 01 Jan 0001 00:00:00 +0000

Quentin Tarantino is not cool.

Mon, 01 Jan 0001 00:00:00 +0000

We have a sense of where we should end up. We want a smoother experience iwth low cognitive load. But where do we start?

How would Mik answer that question? Hiring, make sure you have the people you need to do the job.

There are three levels of devops / platform engineering adoption. Low (Struggling) Medium and High (Succeeding).

There is a big levelling up in terms of returns as you progress the adoption curve.