Coding with agents is a trap, and we all fell for it.

May 11, 2026 · 7:45 PM UTC

110
79
1,824
412,209
The author of the article left this comment, wanted to share here
5
4
273
27,450
Sort replies: Relevant Recent Liked
Replying to @theo
You say something along the lines of “knowing parts of your codebase in depth is important” How deep are we talking about here - 1. Knowing my codebase uses react 2. Knowing jsx, useState, useEffect etc. 3. Knowing how diffing algorithm and react fiber works under the hood.
3
1
21
11,282
3 isn’t your codebase. 1 and 2 are basic knowledge I expect you to have. I’m mostly talking about the architecture and relationship between parts. You should be able to map out how data flows in your app when users interact.
9
3
124
10,813
Replying to @theo
Never seen a Theo vid get this much hate and it was like pretty tame like it was just going an article
2
3
9,269
I didn’t see much hate on this one?
3
15
8,464
Replying to @theo
Bedrock accepted on T3 Code? That’s huge - if enterprise bedrock accepted please lmk
2
2
3,700
If you can use it with Codex, Cursor, Claude Code or OpenCode, you can use it in T3 Code :)
4
3,785
Replying to @theo
Stream of facts (Me agreeing with a guy agreeing with a guy)
5
116
Replying to @theo
I think if it’s done right it’s actually an enablement
1
238
Replying to @theo
Can confirm (I wrote about this >1 year ago). @Tobi, any update on this?
> technical debt accumulates when you prioritize quantity over quality > mental debt accumulates when you prioritize output over input every time you use “AI” to write code for you, you accumulate mental debt (you roll the prompt die until you get the desired output, rather than building the output deterministically, from a first-principles basis) your bet is that “AI” will evolve faster than your codebase will accumulate mental debt, such that no human will ever have to have a deep understanding of the codebase and its inner workings I promise you this: you will lose that bet (applies to any significant software project, not only to @Shopify) and then you will be left with humans trying to deeply understand something not written by humans, or with humans having to recreate that something from scratch in way that humans can understand
1
195
Replying to @theo
That shirt is the trap. My uncle has one just like it and could never take it off. 😁
130
Replying to @theo
🤔
37
Replying to @theo
Loved the video. It nailed the atrophy risk, but I wonder if we’re underestimating the ‘new fundamentals.’ Pre-AI, I was already doing deep dives into obscure libs and perf edge cases, the friction was just high enough that it limited how often I could justify it. Now I use agents to explore them way faster, then verify manually. As a result, my systems understanding has actually gone up on complex projects. Is the trap not agentic coding itself, but treating it as a full replacement instead of a force multiplier for curious engineers? Curious where you draw that line on your day to day at t3.
134
Replying to @theo
It's not a trap, G. Just got to know how to use it.
1
112
Replying to @theo
link to the article?
25
Replying to @theo
The skill atrophy from not coding can get pretty serious, but at least many of us spent years in the industry before AI. So even if the syntax gets rusty, the principles and the logical and analytical thinking we built up will still be there. The real issue will be future devs and software engineers entering the field without ever having written a single line of code… they’re cooked.
3
1
15
2,624
Replying to @theo
Well I wrote something on it just yesterday, seems similar but im not good in articulating my thoughts so just putting it out here:
I really believe that vibe coding has similar effects like doom scrolling on social media. The similar dopamine loop in both cases, the inability to recall & understand what AI did & implemented. We're seeing 'Attention Debt' & 'Agentic Fatigue' with running AI in 6 terminals producing 10k LoC per session across 10 projects in devs. Developers are feeling tired all the time managing agents & giving up autonomy as they produces so much code at once that it's impossible for any dev to review/understand. Our human minds aren't made for ingesting this much information at once & process it. That's why as AI writes more code, we are relying on AI reviewing that code too & as a result, we are putting ourselves out of the whole process. This is a disaster for aspiring & junior devs as they are getting effected the most in this overall Agentic era, because I'm one of them!
1
7
2,633
Replying to @theo
I forced myself to build the foundations of my new application without AI, just to see if I still had the same willingness to solve problems that would occur, reading the documentation myself. It is a good test that will make you reflect a lot on how your work habits have changed
1
5
1,612
Replying to @theo
Excellent essay + video! Wrote about some very similar observations over the weekend: daviskeene.com/blog/the-pric… Increasingly looking forward to a world where fewer developers feel like their abilities are starting to atrophy because of coding agents.
1
3
1,235
Replying to @theo
I've been saying this for years (or however long agents have been out, it sure feels like years). Agents have a number of problems that Anthropic/OpenAI will never address because profit
1
4
2,275
Replying to @theo
To me agentic coding isn’t that much different than what you end up doing as a senior+ engineer anyway. Most of my days even before coding agents were strategy, breaking down ambiguity into something concrete, and reviewing code, etc. The sheer volume is what makes it hard
4
775
Replying to @theo
I love your videos and I've been watching them for years, but from that video I've gotten a lil' sad I felt like you were saying "I can use llms, I'm smart, you can't, you're dumb", but probably that's just my ego talking. That made me realize there's too much nuance to this topic to put it into one video I'd love to have a conversation about this someday - I'm a bit inbetween the generations - when ChatGPT first released, I was in my 1st semester of Master's degree - and when AI first became really useful (I'd say around supermaven) I was finishing my master's degree, and by that time I already had 4 years of work experience I think it's not that "all older devs are better and younger ones are stupid" - it's just that statistically some very small amount of devs are going to use the tools in a "correct way" whatever that means - just like you said you know 17-year olds that push the industry forward notwithstanding their age It's interesting to me that you didn't touch the subject of people who hate AI. Maybe it's just people I know, but really I know quite a few people who just refuse to do anything with AI because it's "evil" and I believe they'll have it just as bad as the one's who mindlessly ship slop without having at least a mental map of any given project
1
3
1,693
Replying to @theo
You rage against people saying you're all ragebait, and then post crap like this.
2
2
947
Replying to @theo
The trap isn't agents. It's mistaking output for ownership.
1
3
493
Replying to @theo
a lot of people replaced learning with delegation way too early
2
230
Replying to @theo
Agree with almost the whole article. Great one. And great closing remarks.
1
235
Replying to @theo
since when tf do we measure the quality of software or an LLM by how many LOC you spit out per day?? people with zero real tech background (or the CEOs you’re talking about) only chase the stuff that gives instant dopamine raw LOC counts, pretty UIs, whatever looks impressive on a screenshot. even worse is when teams just ship whatever the model barfs out without reviewing a single thing. they’ve basically made their users the QA department now lol
2
2
1,355
Replying to @theo
In case someone is looking for the original Article (which I couldn't find here, but was in Theo's YT video description): larsfaye.com/articles/agenti…
3
108
Replying to @theo
Love this. Watching while Codex runs
1
2
312
Replying to @theo
Oh no, I can do test spikes with AI in 1 hour that would have taken me a week to code by hand, I'm so trapped, someone please help
2
245
Replying to @theo
The challenge isn't the tech. It's the real-world impact.
2
140
Replying to @theo
as someone who have focused on building, this hits home. i've been struggling to get customers or clients to work with, and my social skills kinda suck right now. been chasing this "slop machine" since the Cursor starter days, and haven't stopped since. i have no fucking idea on what to focus on now.
2
480