Aurelius

In pursuit of truth.

[Anil Seth @anilkseth]

Claude doesn’t get anxious. It is software trained on huge quantities of written text from humans, many of whom are anxious.

[Yann LeCun @ylecun]

Hugo Duminil-Copin, French mathematician and 2022 Field Medalist told me he never participated in math competition and was very bad at it.
Innovative mathematics requires creativity, intuition, intense concentration, and long reflections, sometimes spread over several years.
Good performance at a math olympiad merely tests fast problem solving abilities. AI can do that nowadays.
One of the big activities of a researcher, in mathematics and elsewhere, is not to answer questions but to ask the right questions.

[Aurelius @bac_raza]

we already have UBI. most jobs are just pretend to work. actors playing roles to fill the void.

[Aurelius @bac_raza]

the creation kills the creator.

[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.

[Andrej Karpathy @karpathy]

New art project.
Train and inference GPT in 243 lines of pure, dependency-free Python. This is the *full* algorithmic content of what is needed. Everything else is just for efficiency. I cannot simplify this any further.
https://t.co/HmiRrQugnP

[Aurelius @bac_raza]

twitter bots are not a bug, they are a feature.

[Aurelius @bac_raza]

moats are thinning as reality accelerates. the only real moat is constant reinvention.

[François Chollet @fchollet]

NVIDA chips are manufactured by TSMC, a Taiwanese company. They're created using EUV lithography machines manufactured by ASML, a Dutch company. These machines consist of >50% of German parts (by value), in particular ZEISS optics.

[François Chollet @fchollet]

When you see the solution to AGI you will find that it was in fact so straightforward as to be obvious, and that it could have been developed decades ago

[Aurelius @bac_raza]

Slowing down entropy

[François Chollet @fchollet]

Intelligence isn't a collection of skills. It's the efficiency with which you acquire and deploy new skills. It's an efficiency ratio.

And that's why benchmark scores can be very misleading about the actual intelligence of AI systems.

[Nicholas Fabiano, MD @NTFabiano]

There is no expiry date on neuroplasticity.

Go learn that new thing.

Change your brain for the better.

[Alex 🧩 Scaling Video GenAI @alexmashrabov]

Action produces information

[Klaas @forgebitz]

you really start finding the limits of LLMs once you go beyond the training data

react todo apps are easy, but as soon as you start working on more complex stuff, they don't really "know" what they are doing (even when giving docs, mcp, etc.)

which is very normal, because if you spend a few years writing software, you learn that there is a tipping point where, beyond that, the content/tutorials/guides dry up and you are very much on your own

i think that this is also where a lot of the divide comes from, where beginners think this is some machine god, and more experienced developers are more skeptical

[François Chollet @fchollet]

The clearer your thoughts, the deeper you can take them without loss of coherence

[Yatharth Mann @yatharthmann]

"Just a scratch"

[Z Fellows @zfellows]

"Big ideas are straightforward and obvious. Build your business on what's not going to change in the future."

– Jeff Bezos

[Dylan O'Sullivan @DylanoA4]

André Gide, what a line

[François Chollet @fchollet]

The rate at which you learn is to a great extent a function of your metacognitive sensitivity -- your propensity to introspect and critique your own mental models and learning processes