The activation vector reducer dilemma
I think I have a cold start problem. Something to do with getting something from zero. I like to maybe start from one or two.
It’s almost like my ideal team semantics is the relay race type of strategy. Someone starts the running. I pick up at number two. I give them a sprint so far ahead, number three had better just keep it up. Or you put me at number three, you know?
I could derisk your loss somewhat. And then the problem, the challenge is life isn’t always a relay race, or not all situations are a relay race. And to be good at relay, you kind of need to be good at running, doing the 100 meters, 200 meters.
However, in situations where I’m not alone and I’m working with a team, there is the cold start problem. It means someone needs to just start, warm up everything, and I get going.
And actually, the more I think about it, the more I see this pattern in other parts of my life. Even when I miss my flights, it’s usually because of how long it takes for me to actually pack my bags. And packing my bags in itself is also sort of a ritual because the question is, how long am I going to stay?
What am I going to do? What do I need for this trip? Which bag do I carry? The big one, the medium sized one, the extra large one? Do I need two medium sized bags?
[inserts GIF with going crazy in my head interpretation]
And usually every time I miss the flight, I could easily just attribute it to how long or how late I packed. Meaning, if I want a high chance of making a flight, I need to have packed five days before because anything 72 hours to my flight is a risk. It’s always a big risk, regardless of how.
And usually also the same when I arrive somewhere new. It takes me about five to seven days to unpack in every new location.
Which is usually why I always have a small bag of utilities that can get me through the five days, and I do not touch the suitcase.
It’s kind of like just the dilemma of ignition. And this is why I don’t do online courses.
And in project teams as well, if I’m inundated with too much information, it could lead to a point where I just break down and wait, aka the cold start problem.
So fixing the activation energy is usually crazy because, I have done a little introspection, and here’s what I think.
I think ambiguity can be one. But the thing is, I actually love ambiguity. It’s my bread and butter. I like environments where a lot of things are uncertain.
But the slight effect of that is ambiguity over a long period of time gives me so many choices. And if I find myself in a situation where there are five possible starting points, ten possible starting points, then, you know, the activation energy just became a lot higher.
And usually, for each starting point, I always like to see the full picture first before I even dig my legs in. And that’s challenging because it takes time and research, and only till I have a view of the full picture do I then dig my legs in.
So every once in a while, it’s like maybe partnership could help, where someone else can analyze, get a view of all the pictures and hit it. But I have just realized that in a team, my greatest value is compounding.
I can take you from 10 to 60, 10 to 80, sometimes 20 to 200. You know, I’m the 10xer, the catalyst.
But if you put me at the starting point, it usually is just tricky. Always tricky! I eventually get through it, but I have seen how the struggle is to get through it compared to situations where you give me:
here’s what we did, here’s what we have, this is the context, this is the destination, and we’re trying to get here.
This is why I excel as an Independent Contributor. I come, not at the beginning of a project, just mid, enough that the individuals in the organization either failed at an attempt or are at a loss of time, so they hand over their scraps to me.
And I’m like, okay. Then I fix the bottleneck, turn things around, usually in record time. So, I’m not sure why I’m writing this. It’s just a little bit of an observation.
Maybe as I observe more, I will document towards it.
Continue Reading
What did the agent do?
I asked an agent to do something. It did it. I kept raising the bar until it ran on its own across sessions I was nev...
Prompt Engineering Manager
I used to write technical vision as Gherkins, C4, OpenAPI specs and let a team of engineers run with it. Now the team...

My Loop Engineering Workflow
Three nested loops run my agent harness: objective, execution, feedback, refined objective. Fork-join in the executio...
