fari @forestfari

AI is making kids dumber. It should be making them geniuses. Introducing Koji, the first AI tutor that gets kids to actually think. 👇

Expectation: the age of the IDE is over Reality: we’re going to need a bigger IDE (imo). It just looks very different because humans now move upwards and program at a higher level - the basic unit of interest is not one file but one agent. It’s still programming.

Andrej Karpathy @karpathy

3 months ago

@nummanali tmux grids are awesome, but i feel a need to have a proper "agent command center" IDE for teams of them, which I could maximize per monitor. E.g. I want to see/hide toggle them, see if any are idle, pop open related tools (e.g. terminal), stats (usage), etc.

300 117 3K 1.4M 1K

A few random notes from claude coding quite a bit last few weeks. Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent. IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits. Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased. Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion. Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage. Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building. Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it. Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements. Questions. A few of the questions on my mind: - What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*. - Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro). - What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music? - How much of society is bottlenecked by digital knowledge work? TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.

same

Thariq @trq212

6 months ago

my favorite way to use Claude Code to build large features is spec based start with a minimal spec or prompt and ask Claude to interview you using the AskUserQuestionTool then make a new session to execute the spec

312 571 8K 2.3M 15K

@brilliantorg @strickinato Big thanks to @avidthunder for his work on the course, too! 🙏 We have the best team 😭

So grateful to work here, where we can actually build the future of education. <3

Brilliant.org @brilliantorg

8 months ago

.@strickinato @forestfari go behind-the-scenes of our new interactive Python course, where breaking things is expected. With smart autocomplete handling syntax, you focus on the real work: logic, iteration, and debugging. By the end, you’ll have built a cybersecurity system,

3 13 34 10K 8

@suekhim 💯

There’s a better way to learn Python. In the era of AI-autocomplete, a missing colon isn’t where beginners should get stuck. With our new Thinking in Python course, you *become the autocomplete*: tap to complete programs, implement correct logic, and actually understand your code and how to structure it. Build security features for a networking app while mastering variables, Boolean logic, loops, and more. Launching next week!

@suekhim Reasons why LLMs write such mid code and it’s easy for beginners to get stuck in vibe-coded nightmares

@suekhim The boundary between math and computer science has always been wobbly (partially because it’s fundamentally nonexistent), but for those that still think in terms of disparate disciplines, it’s worth pointing out that these skills are crucial to all of computer science.

Things mathematicians do, that LLMs still struggle to do: 1/ Mapping a problem to a simpler, isomorphic problem; having novel insights about what problems are isomorphic 2/ Reducing the complexity of the problem to be solved by solving a special case of it first 3/ Checking the result by a variety of mechanisms to ensure that the correct answer is overdetermined 4/ Successfully stepping back and asking: Does this make sense? Did I apply the right procedure the right way? Should the result look like this? These critical thinking skills seem to transfer outside of math, too, especially for those who have learned math deeply enough that they’re reflexive.

We suck at this test and yet we’re crushing the OECD on growth. Teaching isn’t a prestige profession in the US, so far that’s worked out “great” for us, and that’s the problem. The people who would have been absolutely amazing teachers instead being hard at work tagging training data for much higher pay – is that dystopic, rational, or both? Probably both – perhaps it’s rational that topline growth for the US is more driven by how certain talented individuals are utilized, than by the average level of education. But it’s dystopic that it results in incentives and organization that lower the quality of general education and exacerbate inequality. This is also why I find the case for UBI both rational and dystopic. If wealth and power is increasingly concentrated in the 5% who create ~all the GDP growth, average education won’t be a national priority. We can cheer on the LeBron Jameses of capitalism at FAANG, who will win our games. But for individual quality of life and healthy civic participation, it matters a lot. As long as we reduce things to crude optimizations (like GDP) to decide whether better education is a national priority, we’ll continue to have massive disparities in opportunity, income, health, etc.

Chamath Palihapitiya @chamath

9 months ago

We need competition in how our kids are educated - the current monopoly is bankrupting us.

140 164 2K 136K 95

When is noise good? When is slow better than fast? What are the limits to language, & to knowledge? This week on #ComplexityPodcast, @FaritaTasnim (@MIT) & SFI Prof David Wolpert discuss their forthcoming research on the #Thermodynamics of Communication: complexity.simplecast.com/episodes/94

Very strong opening to a math video. Ok, you have my attention.

Fascinating work now presented by @FaritaTasnim at @APSMeetings shows how even simple models models of communication channels exhibit non-trivial thermodynamic properties. meetings.aps.org/Meeting/MAR22/…

Wonderful news! SFI Trustee @B3_MillerValue just gave us the largest donation in the Institute's history — possibly the largest single contribution to complex systems research ever — which he calls "a bet on the future of humanity." 🥳 THANK YOU, BILL! 🎉 santafe.edu/news-center/ne…

Vince Staples brings the sound of his new self-titled album to life for his Tiny Desk (home) concert. n.pr/2UJsbtx

Ecstatic to have started my stint as a visiting researcher at Santa Fe Institute, where I engage every day in deep thought about nonequilibrium statistical physics (to describe the emergent order in living systems) - with David Wolpert! This place is a research dream! @sfiscience

“Neurophysiological evidence for cognitive map formation during sequence learning” - new preprint! biorxiv.org/content/10.110… This was a really fun project, exploring some interesting ideas about how the human brain supports learning abstract latent spaces (1/n)