The AI Great Leap Forward | Hacker News Reader

The AI Great Leap Forward

leehanchung.github.io

142 pointsjodah67 comments

Liked the article in general, but

> These apps will win awards at the next all-hands. In two years they’ll be unmaintainable tech debt some poor soul inherits and rewrites from scratch.

Huge assumption/prediction that I think is actually just wrong. There's this weird assumption from a certain crowd, never justified or explained, that tech debt accrued by AI is now, and will forever be, impossible for AI to address, and will for some reason require humans to fix. Working at pace with agents I accrue tech debt every day, then go through the code nightly, again with agents, to clean and tidy everything up.

The more I see this view espoused the more bizzare it seems. People's assumptions seem to be "if AI couldn't one shot this perfectly the first time, then it's useless to try to have it go back over the codebase and identify and address issues". This doesn't match my personal experience at all, second or third passes over code with CC or Codex are almost always helpful and weed out critical issues, but I'm open to hearing from the rest of the HN crowd on their experiences on this.

Tech debt used here is likely a catch-all term, and you're disagreeing, reasonably so, with one definition.

I think human understanding of the surface area of a company is already very unwieldy. AI balloons the surface area. at some point using more AI to solve AI is reasonable! But to whatever extent a human needs to interface and manage this world, that's the accrued debt.

AIs don’t produce well organized code. They duplicate effort, which is tech debt. Maybe one day they will be able to clear their own tech debt. And who knows, maybe they’ll still be heavily subsidized by VC money then.

This is just false.

Having them clear it is trivial. I have my harness refactor automatically on a steady cadence, something I could never afford to take the time to do manually.

So you have an AI refactor AI generated code? What am I missing here, if AI is the cause of the tech debt because it doesn't write great code, won't you just end up with more tech debt if you ask AI to refactor it?

If a human produces tech debt, do you think a human can't refactor?

Most of the time a human works over code multiple times, and still produces tech debt.

Give an AI agent enough time, by prompting it multiple times, and explicit instructions to look for and address tech debt of various forms, and it will.

Yeah I must be missing something again. Comparing human to AI here seems to be fundamentally wrong. A human will learn over time and improve their mental model of a problem and ability to code. An AI agent for the most part is fixed by its model. I just don't see how pointing an agent at AI generated code to refactor without direct human guidance results in better code.

Maybe you can describe what the various forms of tech debt are that you are talking about?

> Yeah I must be missing something again. Comparing human to AI here seems to be fundamentally wrong. A human will learn over time and improve their mental model of a problem and ability to code. An AI agent for the most part is fixed by its model. I just don't see how pointing an agent at AI generated code to refactor without direct human guidance results in better code.

There is no need to improve their mental model of a problem and ability to code to recognise the refactoring opportunities that already exists in the code. It only takes a sufficient skill, and effort invested on refactoring. The way to get a model to invest that effort is to ask it. As many times as you're willing to.

> Maybe you can describe what the various forms of tech debt are that you are talking about?

Any. Whether or not you need to prompt much to address it depends on consistency. In general I have a simple agent whose instructions are just to look for opportunities to refactor, and do one targeted refactor per run. All the frontier models knows well enough what good looks like that it is unnecessary to give it more than that.

The best way of convincing yourself of this, is to try it. Ask Claude Code or Codex to "Explore the code base and create a plan for one concrete refactor that improves the quality of the code. The plan should include specific steps, as well as a test plan." Repeat as many times as you care to, or if in Claude Code, run /agents and tell Claude Code you want it to create an agent to do that. Then tell it to invoke it however many times you want to try.

But you can just re-roll until you get something good.

A human has taste. They learn over time from codebase patterns and develop a sense of when an abstraction can be reused, improved or refactored. Agents often generate repeated code because the original file wasn't added to the context, and it's up to a human reviewer to recognise this.

In my experience, an agent will rarely recognise a common pattern and lift it into a new abstraction. It requires a human with taste and experience to do it. For example, an agent will happily add a big amount if branches in different places of the codebase where a strategy pattern or enum would be better (depending on the language).

If you have a working prompt or harness that ameliorates this, I'd be glad to see it.

You can organise the code well once, template that and put guardrails in place for it to follow the structure you and the team have agreed is good. The engineering task becomes building the system that is capable of building the system to a high standard.

It will usually just keep on adding workarounds for workarounds. Tech debt increases.

Avoiding this with Opus has been trivial in my experience.

Even with Sonnet or "lower" models like Kimi is trivial. The only thing I still find with AI-generated code is some degree of overengineering.

Skill issue. We don't have that problem. The opposite is true. Every time the harness does something we're not happy with we figure out how to engineer out that failure mode. Tech debt decreases.

so if these "guardrails" really do work (which they don't), why wouldn't you just use them up front?

This also seems to implicitly assume that ai models won't get better - a bet I am not willing to make currently..

Agreed. The confidence people have to predict what these tools will be capable of two years down the line, when it's barely been over a year since Claude Code was first released, is astounding.

Models get better with money (reinvestment).

But if there aren't enough returns soon the money will eventually dry up for OAI and Anthropic and Google will not be trusted with their cash balance.

Its amazing how people here think that money is a play-thing and this dance can go on forever. It cant and wont and the fear-induced marketing doesnt work forever either.

This is a false equivalence. Models get better with more & better data.

Both more data and better data are very expensive. Procuring... Handling... All of the above...

You can spend bottomless piles of cash and by not doing the right things not get there. I can count on one hand the number of times I've seen business/investor incentives line up with r&d incentives.

There's no guarantee that there is enough or good-enough data, regardless of how much money you have.

They're not getting better data. They're now ingesting their own generated slop and bad habits are being reinforced.

Agreed, but it's a bit nuanced. I'm working on a fairly complex project now in a domain where I have no technical experience. The first iteration of the project was complete garbage, but it was garbage mainly because I asked for things to be done and never asked HOW it should be done. Result? Complete, utter garbage. It kinda, sorta worked, but again, I would never use it in anything important.

Then we went through ~10 complete rewrites based on the learnings from previous attempts. As we went through these iterations, I became much more knowledgeable of the domain - because I saw failure points, I read the resulting code and because I asked the right questions.

Without AI, I would likely have given up after iteration 2, and certainly would not do 10 iterations.

So the nuance here is that iterating and throwing away the entire thing is going to become much cheaper, but not without an engineer being in the loop, asking the right questions.

Note: each iteration went through dual reviews of codex and opus at each phase with every finding fixed and review saying everything is perfect, the best thing on earth.

I'm seeing similar process but on large teams still finding this output to be unmaintainable.

The problem is that vanishingly few people actually understand the code and are asking the agents to do all of the interpretation and reasoning for them.

This code that you've built is only maintainable for as long as you are still around at the company to work on it -- it's essentially a codebase that you're the only domain expert in. That's not a good outcome for companies either.

My prediction is that the companies that learn this lesson are the ones that are going to stick around. LLMs won't be in wide use for features but for throwaway busy-work type problems that eat lots of human resources and can't be ignored.

I left my last company job just before "AI-first engineering" became mainstream, and you confirmed what I was feeling all this time - I have absolutely zero idea how teams actually manage to collaborate with LLM-managed projects. All the projects that I'm working now are my own and the only reason why I could do this is because I had unlimited time and unlimited freedom. There's no chance I would be able to do this in a team setting.

I'm positive that the last company's CEO probably mandates by now that nobody must write a single line of code by hand and there's likely some rigid process everyone has to follow.

Fun times ahead.

I agree and commiserate. In the near term my picture is pretty grim. There's fantastic uses for these tools but they're being abused.

I was big on correctness, software safety (think medical devices, not memory) and formal proofs anyway, so I think I'm just going to take the pay cut and start selecting for those types of jobs. Your run of the mill SaaS or open source+commercial companies are all becoming a death march.

> Your run of the mill SaaS or open source+commercial companies are all becoming a death march.

Most of them already were death marches to begin with, now they are firing squads

I agree with this.

If anything there will be less tech debt, because it's easier than ever to clean up.

Unit and end-to-end tests are free now.

Telling an agent to go through your code base and find bugs is free now.

Telling an agent to generate all possible user interaction to your API and then performance test it is free now.

Everything is so easy and takes much less time than it used to.

> Unit and end-to-end tests are free now.

Explain please. I might agree with discounted.

> Unit and end-to-end tests are free now.

Not free because agents have a big tendency towards useless tests, so you need to verify them and make sure they're testing real things and the things that matter. I'll agree that it's a lot easier to generate thorough tests matching a spec, though.

I have some recent experience with this, and it's quite quantifiable. I helped two people with no UI experience get their AI generated project deployed.

It probably won't pass a security review (It rather curiously generated a document.write call in a React project to print a component). I had to disable a number of accessibility related lint checks to get the code committed, as it didn't pass them. It has no tests.

The effort to meet the company's stated standards is genuinely likely to be greater than creating the thing.

The organisation where my wife works ordered all mid-level leaders and above to take a mandatory AI course, mostly remote with two days on-site to present their capstone projects, costing 2x K USD (not including flights). The capstone projects sounded impressive, and the course was celebrated as a resounding success. However, one year later, as far as I know, none of the capstone projects have been implemented, including my wife's and other ones I know of.

Having looked at some of the project descriptions, I realised that they would need to invest far more manpower, special expertise and time if they wanted to implement them with a moderate chance of success.

I believe this is not uncommon in large organisations worldwide.

BTW, it’s great that somebody has drawn a comparison with China’s Great Leap Forward. Not many people know about it and it always serves as a stark reminder of how crazy state-ordered “progress” could be.

Decent sentiment and analogy, but writing this with AI with hackneyed examples undercuts the point

I’m noticing one hallmark of blog posts made by people who talk to LLMs all day: they have 1-3 interesting points hidden in paragraphs upon paragraphs beating the horse dead. Your favorite LLM might tell you every thought is brilliant and all your words are beautiful, but please… edit it down. At the very least, out of respect for other people’s time.

Haven’t you heard? Putting in effort is not cool any more. The best they can do is ask an LLM to edit it down.

It's called body text or even "bread text" in some languages. It was historically meant to pad the pay for bread (writers got paid per word). Americans still do to this day and writing and blogs reflect it as well.

I didn't even realize that this is (allegedly) written with AI. If it's AI, then it's the kind of AI writing that's closer to the real deal and that I want to see more widely applied if AI is to be used.

I actually found the examples to be both on theme and reflective of "If history doesn't repeat, it certainly echoes".

Note that the author of this blog post is also the author of a soon-to-be-published Manning book on safely implementing AI systems.

If you want to show that that there's a risk of disaster you need to do better than making a silly analogy. Companies will often start expensive projects that fail and then they pick themselves up and move on. Big, profitable companies can afford bigger failures. Google has had a slew of failed projects, and Meta's metaverse stuff tanked, and they're still fine. They can afford to experiment.

So which companies are betting so big that it might actually threaten them? Oracle maybe?

Seems clear to me that OpenAI at this point is a Ponzi scheme waiting to collapse. This is why they are trying to IPO and dump their shares on the public market before they go bankrupt.

Suppose they do somehow collapse. How does that cause wider problems? Their competitors will pick up customers.

If they collapse, then because their value proposition doesn’t add up. It’s unclear why that should be different with their competitors then.

It looks like nobody is collapsing, but OpenAI might be behind Anthropic now:

https://www.axios.com/2026/03/18/ai-enterprise-revenue-anthr...

https://x.com/albrgr/status/2041288324464451617

"Google has had a slew of failed projects, and Meta's metaverse stuff tanked, and they're still fine. They can afford to experiment."

Only with the blessing of shareholders. Frankly Google's search box and ad-tech has been carrying all of its failed bets but at some point people will start questioning if Google is returning enough cash given the results of new investments. Google's management does not own the cash - it holds the cash on behalf of the owners.

Which shareholders do you mean? Mark Zuckerberg holds >50% of voting rights for Facebook. Sergey Brin and Larry Page hold >50% of voting rights for Google. That means management gets to do what it wants, within very broad legal limits.

On the other hand, how the stock does will matter to other employees because they’re shareholders and they have a stake in the outcome.

At work, some team are going AI full speak ahead, for a year now

Their product still sucks, their roadmap is still a joke, their product is still buggy, they are more people than ever and their cost has increased

  what did any of this actually produce?

> Today’s backyard AI looks like AI. It is not AI.

Getting real tired of people new to AI thinking only recent LLMs are AI somehow. BoW was a pretty solid technique and that only requires you to learn how to count to one.

We can thank our AI overlords like sama and damodei for that.

>A prompt template behind a REST endpoint is not a model.

Not pulling any punches over there. It does feel like 95% of the "AI industry" consists of wrappers and associated tools.

Great rant! Claw based propaganda posters makes me smile.

Oh god, don't get me started on this. The article goes full opera-level tragedy, like we're all marching into some corporate gulag where AI eats our souls and the lights go out forever. "The famine comes later" my ass. It's peak doomer porn, written to make you feel like the sky is falling instead of just another round of executive circle jerking.

The corporate world has always been 80% lies, fake KPIs and theatre. "Synergies", "disruptive innovation" "digital transformation", same shit since the 90s. Managers don't give a flying fuck about your clever moat. They wake up one day, get a spreadsheet from McKinsey saying "cut 15%" and boom - your undocumented wizardry gets deleted along with your badge. Nothing personal, just Excel doing what Excel does.

Yes, the corporate bullshitry has been turbocharged with AI now. But it's nothing new and nothing that much tragic. At the very least the same AI can help me finally release personal projects that have been collecting dust for years. Who knows what the future will bring. I'd be much more worried of oil supply chokehold than of AI turbo circus in the corporate world. No oil means not having enough food tomorrow; or medical supplies. My child might die because of this. But AI temporarily causing perturbations at work is just another round of corporate theatre. Been there many times.

Employment danger is real, but not apocalyptic. Some jobs will evaporate, sure. But even as the same articles states, now once thing ("AI know-how") replaced another thing ("domain knowledge siloing"). The corporate machine still needs warm bodies for the messy human parts: sales, talking to customers (customers hate talking to a robot, what a fucking surprise), covering ass. I would say, covering ass is the most important one, along with delegating the project management to someone else below on the corporate hierarchy, so upper management wouldn't have to work and would only keep asking for status updates. They would always need someone to type the actual AI requests. It's not like top management or VP would ever do that, neither they would ever run it automatically, since AI can delete production (happened many times), and they don't want to be the scapegoats.

So yeah, the article is overdramatic trash for clicks. AI is just another round of that circus. The "famine" won't be real, it'll be a bunch of overpromises, just as usual. Same as it ever has been.

>"Synergies", "disruptive innovation" "digital transformation", same shit since the 90s. Managers don't give a flying fuck about your clever moat. They wake up one day, get a spreadsheet from McKinsey saying "cut 15%" and boom - your undocumented wizardry gets deleted along with your badge. Nothing personal, just Excel doing what Excel does.

The buzzwords you cite are the vulnerabilities of the corporations which predator consultancies rely on to make sales. I don't know that the corporate world is 'about' those things so much as it suffers from them.

The outputs were wrong 2 years ago maybe.

Try multithreaded code, it's wild (how wrong it is). I wonder how/where next models will learn thread safety.

Good article and it's on the mark

What’s the story with Klarna? Any details around it?

It’s the punchline at the very end of the article. They ended up with a different SaaS vendor.

Yeah I read through it but all of that is surface level. Any real insider info?

Not sure why I was downvoted. I read the post and the linked articles.

They got rid of many employees replacing it with A.I thinking it would benefit the company. Now they’ve had multiple security issues, are under investor scrutiny and can’t rely on a robot to get them out of real world predicaments.

don’t be fooled, they lie and deceive too. Klarna who? Investor fraud its what they do. #shutdown

Yeah, people being able to access accounts via recycled phone numbers is not good (this wasn’t their first time either lol and they’ve been fined for similar stuff before by Swedish regulators)

Then the KYC bypass allowing fraud rings to buy 20k worth of products with stolen identities is outrages.

Then the IPO stuff… That entire thing is a mess and there’s so many issues with their F-1A and being dishonest.

Though this is just my subjective opinion and shouldn’t be taken seriously.

The AI isn't a breakthrough; it's a distraction from the fact that the company's security and finances are a mess.

It’s quite funny to now see Klarna offloading their security issues onto other companies (Google) while simultaneously suing them for $8 billion (PriceRunner v.s Google).

Yeah true, mid 2025 they announced it as “A.I innovation” and “collaboration” to fix AP2 without announcing the company then came back late 2025 to say it’s a collaboration with “Google” for security payment protocols etc — While also simultaneously suing Google for 8 billion with their company Price Runner lol.

I guess they were having security issues for a while now and a few have been publicized recently which are the ones you’re talking about, but they kinda fit into the same issues they’ve had in the past and have been fined for like said by @upsetpotato

Comparing AI to steel production in the Great Leap Forward seems unfair. It's not some communist plan - it's a capitalist free for all similar to the industrial revolutions in the UK/US. It won't lead to a famine, it'll lead to the chaotic creative destruction capitalism usually produces.

You're mistaking communism/capitalism as economic systems for communism/capitalism as organizational structures. The latter is what the argument centers around.

It's been argued frequently that families and tech companies are structured like socialist states. Central planning, flatter structures, division of labor...I'm not starting down that thread or opening up that debate.

This only is not a capitalist structure but capitalism itself doesn't really offer any ideas about structure or governance beyond encouraging the free movement of capital.

Fair enough.

This is a great comparison. The US dominated software industry is centrally planned and in many ways run like a communist country, taking into account the whims of the current chairman in Washington.

If the chairman dictates DEI, DEI it is. Most software developers put up the proper flags in their Twitter "bios" and purged opponents. The same developers now queue to work for Zuckerberg's "male energy" company.

If the chairman and the industry dictate AI, AI it is. The same people who said girls and coal miners have to code now talk about efficiency, products and rationalize layoffs.

This is the product of an industry that has been dominated by bullshitters for at least two decades.

Reply on news.ycombinator.com