Steven Van Vaerenbergh

Researcher and lecturer at @unican. Occasional curator of lists.

[Steven Van Vaerenbergh @steven2358]

Which human bottleneck becomes dominant when cognition is cheap?

[Steven Van Vaerenbergh @steven2358]

@FigureLabs_AI One of my students showed me your product. It looks good—I've added it to my list at https://t.co/D2Sr9R4PUQ

[Steven Van Vaerenbergh @steven2358]

@karpathy Excellent. Also, this is what formulating algorithms in pseudocode has always been about.

[Andrej Karpathy @karpathy]

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. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow.

Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. 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. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. 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.

It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). 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.

[Steven Van Vaerenbergh @steven2358]

In hindsight, we could have seen this coming. #moltbook

[Steven Van Vaerenbergh @steven2358]

@_GunRunn3r_ @4HLsartorialist The whole soundtrack is phenomenal.

[Steven Van Vaerenbergh @steven2358]

@calcsam BOOK

[Steven Van Vaerenbergh @steven2358]

@Jindo_1307979 And the classical one (no exponents):

811051

[Steven Van Vaerenbergh @steven2358]

@Jindo_1307979 Another option:

60E8

[Steven Van Vaerenbergh @steven2358]

@calcsam BOOK

[Steven Van Vaerenbergh @steven2358]

@giffmana This should be required reading for all PhD students in the field.

[Steven Van Vaerenbergh @steven2358]

@calcsam BOOK

[Steven Van Vaerenbergh @steven2358]

“When unlimited power meets limited understanding.” A very decent read on how AI is detrimental to junior coders’ skills. The argument generalizes well to other fields: AI access should come *after* demonstrating mastery. https://t.co/94aGD1zjFq

[Luis J. Rodríguez Muñiz @ljrguezmuniz]

Hoy en @elperiodico hablando de Educación Matemática con motivo del inminente @C2EM2025

https://t.co/JEvqFLejtD

[Steven Van Vaerenbergh @steven2358]

@aphixx @elonmusk Does this work? https://t.co/WHhJjQWTuF

[Steven Van Vaerenbergh @steven2358]

At some point, we’ll have true "vibe mathematics": you describe a problem in natural language, and the AI interprets, rigorously solves, and explains it. But current LLMs still hallucinate steps, so we’re not quite there yet.

[Steven Van Vaerenbergh @steven2358]

In maths, "vibe calculating" has been around for a while. You write the operations and get the result from the calculator, without worrying about their correctness. Compared to LLMs, the correctness is pretty much guaranteed here.

[Andrej Karpathy @karpathy]

There's a new kind of coding I call "vibe coding", where you fully give in to the vibes, embrace exponentials, and forget that the code even exists. It's possible because the LLMs (e.g. Cursor Composer w Sonnet) are getting too good. Also I just talk to Composer with SuperWhisper so I barely even touch the keyboard. I ask for the dumbest things like "decrease the padding on the sidebar by half" because I'm too lazy to find it. I "Accept All" always, I don't read the diffs anymore. When I get error messages I just copy paste them in with no comment, usually that fixes it. The code grows beyond my usual comprehension, I'd have to really read through it for a while. Sometimes the LLMs can't fix a bug so I just work around it or ask for random changes until it goes away. It's not too bad for throwaway weekend projects, but still quite amusing. I'm building a project or webapp, but it's not really coding - I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works.

[Steven Van Vaerenbergh @steven2358]

@AndrewHallFF @KushDesai47 So it may seem. https://t.co/BNBKXGkR1A

[rohit @krishnanrohit]

This might be the first large-scale application of AI technology to geopolitics.. 4o, o3 high, Gemini 2.5 pro, Claude 3.7, Grok all give the same answer to the question on how to impose tariffs easily.

[Steven Van Vaerenbergh @steven2358]

4o image generation is wild. But... that's not a prism, it's a pyramid. GenAI still doesn't get geometry. Is that on the roadmap, @OpenAI? https://t.co/ftxCaEDbcd