Links

Programming is becoming unrecognizable. Karpathy draws attention again to the famous November-December 2025 inflection point.

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December.

He was able to get his agent (most likely his claw) to build a local video analysis dashboard in ~30 minutes

I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. [...] As a result, programming is becoming unrecognizable.

He says that the "biggest prize" is being able to set up Claws well enough that it can handle multiple Code instances for you

The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now.

I am currently still operating at the LLM agent layer. I did setup Jacho but given that his brain is powered by GLM 4.7, he isn't that smart of a claw. I'm considering birthing a new claw and plugging its brain to Gemini soon. I'll be more meticulous this time with its setup so that it can "productively manage multiple parallel Code instances" for me.

He points out that while its not perfect yet, the key is figuring out which parts to delegate to AI to be fully autonomous and which ones to be more involved in.

The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

at the top tiers, deep technical expertise may be even more of a multiplier than before because of the added leverage.

Remote Control in Claude Code. Sometimes when I'm in the zone but I have to step out for a while, I'd really wish I could control the Claude Code session from my phone. There is an official way to do this using Claude Code on the web but it requires a GitHub repo setup and runs on the cloud.

It's also possible to setup Tmux + Termius + Tailscale or use the Happy Coder app which seems to be an open source option which is closest to Remote Control.

Claude Code has native support for this now. It allows us to continue a session remotely from our phone using the new Remote Control feature.

Anthropic is making it really easy for Claudoholics like me to have 24/7 access to our drug.

The Only Moat Left Is Money. Found this interesting post on Hacker News and immediately thought of my blog post from May 2025: Thinking is a commodity. At the end of that post, I hint at writing another post answering the question if money matters now more than ever:

LLM thinking comes at a price, and it can think deeply if you pay more. If you do the “boring” work yourself, you fall behind. Does money matter now more than ever? Food for thought (no pun intended)

I think this post by Elliot captures my feelings very well. Here are some quoteworthy snippets from it:

The effort is gone. Effort was the filter.

When creation was hard, skill was the differentiator: you had to actually be good to make something worth showing. Now the barrier is near zero, so you need reach. Reach costs money or it costs years. Probably both.

He thinks that there's a real chance that those without existing reach are locked out

I don't know if we've already crossed a singularity on this, a point past which new entrants without existing reach or capital to buy it are effectively locked out. I think there's a real chance we have. The uncomfortable version: if you're not already moving, you might never take off.

Something Big Is Happening. Highly recommended reading. It might seem a bit lengthy but definitely worth it.

The author's intention is to invoke preparedness rather than just fearmongering. I can confirm I had a similar feeling about the newer models being so good that they can do the same job as me, much faster and better. I will say, though, that getting them to be that good is a skill in itself.

Here are the key highlights from the essay:

1. "The gap between public perception and current reality is now enormous, and that gap is dangerous because it's preventing people from preparing". This is true, most people in my circle have little to no idea about the rapid improvement in AI in the last few months. As the author says, this is mostly because of their prior experience from 2023/2024 or because they use the free version.

2. "I think the honest answer is that nothing that can be done on a computer is safe in the medium term. If your job happens on a screen (if the core of what you do is reading, writing, analyzing, deciding, communicating through a keyboard) then AI is coming for significant parts of it. The timeline isn't "someday." It's already started". I agree, especially because he says "... coming for significant parts of it".

3. "The person who walks into a meeting and says "I used AI to do this analysis in an hour instead of three days" is going to be the most valuable person in the room. Not eventually. Right now". Makes a lot of sense. The ability to use AI to its full potential is valuable. The bottleneck is not the AI itself; it is the uninformed human with a lack of imagination.

4. "Make a habit of experimenting. Try new things even when the current thing is working. Get comfortable being a beginner repeatedly. That adaptability is the closest thing to a durable advantage that exists right now". This is also the best way to develop the valuable skill of using AI to its full potential.

And this is the most important takeaway:

"Here's a simple commitment that will put you ahead of almost everyone: spend one hour a day experimenting with AI. Not passively reading about it. Using it. Every day, try to get it to do something new... something you haven't tried before, something you're not sure it can handle. Try a new tool. Give it a harder problem. One hour a day, every day. If you do this for the next six months, you will understand what's coming better than 99% of the people around you. That's not an exaggeration. Almost nobody is doing this right now. The bar is on the floor."

This commitment is something I made to myself at the beginning of the year (long before this essay was published). I began to feel a sense of urgency but not the FOMO kind. I was okay with having this feeling because it would push me to use all these tools, experiment with new models and keep up to date with everything happening in AI. The goal is to become a power user of AI.

Maybe you're not Actually Trying. I find myself going back to read this incredible article by Cate Hall every time I feel like something is not going right in my life or that I have disappointed myself.

It's about how even though we might think we are trying hard to solve a problem in our lives, we might not actually be trying.

[...] feeling of effort doesn’t mean that you’re Actually Trying.

These are the key points from that article:

1. People are not just high-agency or low-agency in a global sense, across their entire lives. Instead, people are selectively agentic.
2. It seems like, by default, you are stuck with whatever level of resourcefulness you brought to a problem the first time you encountered it and failed to fix it.

I would be careful when applying the 2nd point to past situations because it is tempting to evade accountability by blaming circumstances in life at that point of time. Instead, I'd tackle it by asking myself:

1. Have I done my best to come up with a set of potential solutions, using all the resources I have?
2. Am I doing as well by myself as I would by any friend who came to me with the same problem? How do I know I’m Actually Trying?

I really like this because it reminds me of the Sam Harris quote:

"On one level, wisdom is nothing more profound than an ability to follow one’s own advice."