Notes

I have been getting into Cate Hall’s book [4], You Can Just Do Things, and I appreciate the discussion on agentic foundation, but it’s just that she means the other agentic than the agentic engineering everyone else is talking about these days.. For Cate, being agentic is from the human perspective about your agency. There is this song, by this dutch band De Staat, “Make Way For the Passenger”, [1] where the music video is a POV of a motorcyclist, riding alongside a few other motorcycling buddies. Then the music crescendos and we see the POV rider gets t-boned by a truck, and wipes out. Then the video rewinds to the beginning, while the music keeps its original forward orientation. The lyrics now are “live fast, die young. THis ride will ride on. This life will die on.” And now, the video direction goes forwards again, and the crescendo plus crash sequence is replayed right until the impact when the screen goes dark. ...

No One Knows Fun anecdote, watching the AI Doc, how I became an Apocoloprimist. (Which btw really is in the samw spirt of film titles for films as Dr. Strangelove or: How I Learned to Stop Worrying and Love the Bomb (1964), directed by Stanley Kubrick and perhaps To Wong Foo, Thanks for Everything! Julie Newmar ). The Director slash interviewer asks everyone what will happen. There is the Existential Risk tribe and the Accelerationist tribe. Also the movie is sponsored by Tristan Harris who is bipartisan here. (Raskin as well). ...

Is decisioncrime the new thoughtcrime? It seemed to him that it was only now, when he had begun to be able to formulate his thoughts, that he had taken the decisive step. This above is the part of 1984 when Winston Smith begins again to reenter thohght. Decisioncrime I find myself –like many others Im sure– deciding less, by outsourcing them to a SlopBot. I must have written almost these very words many times already here, but now I have a new word for the phenomenon of not being allowed to decide. ...

Hmmm There can often be overwhelm in a day to day, that prevents us from enjoying doing work even if we actually like the work. The secret out-of-office test. I have realized a kind of hack long ago that if i toke a day off, i could open up my computer and work at a more relaxed pace and derive way more joy joy than if other people knew i was working. ...

The floor in our kitchen started getting swollen. Hmm maybe it’s the humidity, I thought? So I turned off this extra humidifier I had running. Okay and I put our big box fan on that too. Another day I wonder that okay if it is next to the dishwasher maybe there is a leak? I unscrew the dish washer floor plate and I see yea it is wet there. I am gaffing around with the insulation ard I realize I can just take it out. And then I see it, wow a gentle water leak at the place that connects to the fresh water. I try to capture the water in different ways. First I use an oven pan that is too big because the floor board has now started to open up once I took out that floor plate, because it was putting pressure on the floor. Okay so I use a smaller oven pan but that is still not ideal because when I try to take it pick it up it is really easy to spill anyway I go between that and little dishes that fill up way too fast. I’m trying to work too so I don’t have time to solve this problem since I’m trying to solve work problems too! I get a spare minute and as I’m realizing it is getting late and my building maintenance request is not going to get looked into today I realize there is a good chance I will still be switching this leak oven pan through the night if I don’t shut the water off. ...

Going back to [1], from writing [2], i think in the conversation between DHH and Lex Fridman, they identify a sweet spot for swarms. Safe side So never using parallel agents, on the one hand, is playing it safe. You work iteratively with one agent on the task at hand. The premise is you likely dont know exactly ehat you want or necessarily what you need, at least when youre in the fog of war. For sure you have your own plan for your project, half in your head, half in markdown and a third half you discussed with others and proceeded to forget 😭. But this is just a plan for your sanity, and would be useless to give to your agent because that would be a fool’s errand. And any way, you are still multitasking, becouse you ore still juggling many commitments, but you are not using multiple harnesses within the scope of any one of them. ...

Joel spolsky’s book , [1], his blog, his phrase, is specifically two parts. Intriguingly, just like with students and with large language models, we are more obsessed with the first. In an interesting twist maybe we can slice this further. Nassim Taleb said [6] he valued erudition over intelligence A nature vs nurture thing? Side note, listening to Kathryn Paige Harden [10], describe it, heritability is sort of the expression of a phenotype, and within the behoviorol sense, this is very much tangled up in the nurture. So the expression happens in an environment, like a seed can end up as a stubby tree or a broad tree depending on whether the rain and sun are to its preference. (Anecdotally there was a very very sad orange tree in my apartment which got no direct sunlight and did not last long). ...

A friend sent a Nicholas Taleb quote, that got me thinking, “They are born, then put in a box; they go home to live in a box; they study by ticking boxes; they go to what is called “work” in a box, where they sit in their cubicle box; they drive to the grocery store in a box to buy food in a box; they go to the gym in a box to sit in a box; they talk about thinking “outside the box”; and when they die they are put in a box. All boxes, Euclidian, geometrically smooth boxes.” ...

I suspect Matt Pocock [1] , (along with Dax Raad and Peter Steinberger), is someone who sits kind of in the middle of the debate about just how much more productive coding agents can make you wrt the software development lifecycle. On one extreme is Steve yegge who is kind of declaring orchestration is here, and you need not, ever look at code again, just express your intent and the swarms[4] will manifest it. Or something [5]. Chaos theory be damned. 100x to 1000x productivity is claimed. (Holding my breath?) There are many others in this camp. I would like to understand them better. ...

Exploring Kelley’s benevolent noncompliance [1], has been more formally studied as goal displacement [10], exploration vs exploitation, but also creative deviance. Kelley’s take is about how organizations are kind of like functioning alcoholics, formal and goal-oriented looking on the outside, but held together by unsanctioned thankless sustainment and stewardship of the few. This paper [2] formally explores this topic–no pun intended 😆–and my initial thought is that cultivating deviant creativity is a kind of artform. Helicopter management can likely only will it into wilting. ...