Replit CEO on AI breakthroughs: ‘We don’t care about professional coders anymore’

Should be a nice salary boost for developers in a year or two when all these companies desperately need to rehire to fix whatever AI slop mess they have created.

And I hope every developer demands 2x their current salary if they are tasked with re-engineering that crap.

Like when they stopped trying to Outsource to India or other places where the labor was extremely cheap. But all the code that came back from it was useless and had to be Rewritten by the remaining software engineers still at the company anyway. They'll never learn. They'll simply find a new anti-worker efficiency to Chase.
- But during a shory period of time,managers could force workers to stay late and overwork themselves without pushback.
  
  /s
I'd go further and demand that the team I'm hired for re-write the app completely and not just re-engineer it from the AI slop codebase.
- tbh that's probably going to be cheaper
Bold of you to think there won't be a recession

Bit of a rambly story, but I swear it is relevant.

So previously I worked as a consultant for a company that manufactured a relatively small number of high value (tens of thousands of dollars each) Gizmos in a lightly regulated industry - the requirements weren't too crazy, basically that everything has a serial number and they can prove that any given serial passed the full range of tests before it left the factory. Pretty much the sort of thing you'd want to have if you gave a crap about quality products anyway.

Initially they were using Excel to keep track of this - they manufactured 10 units a week, it worked well enough. Eventually, they got more successful and needed to scale up to 50 units a week, and it was decided that they needed A System to keep track of testing and manufacturing. Their head of manufacturing "looked around and couldn't find anything off the shelf that was suitable" (ie, cost $0, and perfectly matched his aesthetic tastes; mistake #1), so they decided to build their own system.

They had a few in house developers, but they were focused on building new features (things that drive sales, unlike maintaining their reputation for delivering reliable products), so head of manufacturing decided to get one of the production line techs (who was "good with computers" by virtue of having built the Excel system, but was not a software developer mistake #2) to do it.

Eventually, they decided to use Microsoft PowerApps to build the new system - for those with the good fortune never to have seen PowerApps, it's essentially a "no code required" drag and drop UI tool that you script using Excel formulas. Think Visual BASIC or Scratch, but Cloudy.

On the surface this made sense - the developer was proficient in Excel, so use what you know. Unfortunately, PowerApps is designed to rapidly build throwaway UIs over simple data models and lacks some of the things that actual software developers would have thought to ask about:

It lacks real version control - you can "undo" a deploy, but there is no way to discover what changed between versions, or do branches, or code review
Because you can't effectively manage changes to the system, you can't do pre-production releases
Its native database system doesn't do referential integrity
There is no straightforward way to do any kind of locking - and because there is no referential integrity, it's really easy for concurrent users to really mess up the data
There is no way to do automated testing
The development group could have actually documented how stuff worked, requirements, specs etc but didn't, so any time there was any issue you had to play the game of "is this a bug or bad design?"

Eventually, these chickens came home to roost in the form of a defect that slipped through testing that they then couldn't isolate to a particular batch because none of their testing data could be trusted. I was brought in to try and unpick this mess and advise on a replacement system, but between the cost to fix the issue and the lost sales from it they ended up in a pretty bad spot financially and ended up being acquired by an investment group.

Anyway, the takeaway from this is that you disregard experience and judgement at your own peril, up front savings generally don't manifest in the long term and I expect there is going to be a thriving market of consultants brought in to point and laugh at companies that decided that a bunch of cheap, inexperienced developers and a magic talking parrot would build better software than cheap, inexperienced developers being guided and upskilled by an experienced senior developer

This has happened before. GUI tools were going to mean less developers with less cost, but it didn't materialize. Higher level languages were going to cause mass layoffs but it didn't really materialize. Tools like WordPress were going to put web developers out of business, but it didn't really. Sitebuilders like Wix were going to do it, too, but they really haven't.

These tools perform well at the starter end, but terribly at the larger or enterprise end. Current AI is like that. It can help better than I think people on here give it credit for, but it can't replace. At best, it simply produces things with bugs, or that doesn't quite work. At worst, it appears to work but is riddled with problems.

I genuinely believe AI isn't over hyped in the long run. We're going to need solutions to fix our current way of work. But I feel confident it's still further away than the people investing in it think it is, and they're going to be paying big for that mistake.
- it simply produces things with bugs
  
  And the small companies will realize that you need to start investing in your QA again - because NO ONE does adequate QA any more. Every piece of software I interact with these days is shit. If you are a developer, I'm talking to you: your code might be great, but the overall product is shit because no one is adequately testing the functionality. I get it, Testing is expensive... but now that you need fewer developers, hire some QA engineers. Just because we've moved on to 'continuous improvement' doesn't mean we can't do better.
I enjoyed your history. There is a lot to learn about. Thank you.
- If you like stories like that, you'll enjoy the DailyWTF. It's littered with cautionary tales like this.
The last paragraph in this story rings true to me.

I used to work at a company that wasn’t in the software development or tech industry, but wanted to start working on a new software development project. They were willing to put a bunch of money into patents, registering a new company, etc for this. They were unwilling to hire serious experience and expertise—in other words they exclusively hired interns.

I kept advocating to bring just 1 experienced developer to help mentor the rest of the team. But they kept ignoring my advice. They continued spending a bunch of money to make the project that was being worked on look amazing to other people. They didn’t do anything to provide an environment to ensure the project was actually good.

Long story short, I left and from what I can tell the project is not in a healthy state. They have reached the sunk cost fallacy as it has been years since the project was started. There was massive shortsightedness and the person leading the project is one of those “know-it-all” types who doesn’t know jack about software development.

They focus on short term costs and in the long run it costed them more—and not just financially.
Maybe a dumb question but isn’t Power App’s “native database” a SharePoint list? Which has integrity protections because it’s a mysql implementation?
That is something that looks like it could be integrated into ERP system, and they probably already had one sooo seems like a fairly simple problem with a fairly simple solution ( and depending on the size of the company they already might have had either oracle ERP or SAP , unless they were really small ).

As a developer I always enjoy asking these clowns why anybody should buy their products when AI will soon allow consumers to build these apps themselves, which is just a logical progression if you don't need coders to create software.

I know of course that this isn't going to happen anytime soon. It's wishful thinking from their part and it shows a complete lack of understanding both what LLMs can and cannot do and what it takes to design and implement anything bigger than a batch script.

Edit: I also can't help but feel personally offended whenever some corporate drone gloats about replacing developers. It didn't happen with low-code, it didn't happen with no-code and it won't happen with AI. But it hurts every time.

I mean, project managers have been not caring about professional coders since way before the current LLM hype.
I feel like this could be (opinion) the reason why Devin is trying to charge $500/mo for their tool. They know they only have a limited time window until a general-purpose agent from OpenAI/Anthropic/Google/... can directly do everything their product does. So they have to make their money while that gap in capabilities still exists.
- My take: It's either not going to happen at all (which is what I would place my bets on tbh) or it's going to be something that only megacorps like Google or Microsoft can offer due to the enormous requirements of such a system, which would make most of those AI companies redundant, not just their devs.

This clickbait headline has been making the rounds for a few days now. Replit's CEO is not saying that AI has "replaced" professional coders, he's talking about their company's target market.

It's like a website provider making tools to simplify website creation for small businesses so that any mom-and-pop store can have a basic website, and saying "we're not aiming these tools at professional website authors." They're simply not trying to occupy that niche.

Get ready for a bunch of unskilled people making the shittiest apps imaginable.

Those apps have their place. Why shouldn't an "unskilled" person be able to make some little tool that does some specific task they need done? I'm a professional coder and I make "shitty little apps" all the time for throwaway tasks. I think it'll be empowering for the average user to be able to do that sort of thing too.

Obviously, don't go buying such apps and installing them on your own phone or whatever. That's where professionals still have their place.

Why shouldn’t an “unskilled” person be able to make some little tool that does some specific task they need done?

The problem isn't that they're going to use it for themselves. The problem is they're going to either try to make money with it or do some black hat shit that this will help facilitate.

(Also, saying to us that it's obvious not to buy it doesn't really matter considering the huge number of people out there with smartphones and no idea how any of it works.)
- The problem is they're going to either try to make money with it or do some black hat shit that this will help facilitate.
  
  And then they will fail at it, because that's not what these tools are for. I don't see why this is a problem.
  
  If someone is asking you "hey, I want to use this Replit thing to build a competitor to Amazon, I have an MBA so I'm sure I can do it. Want to invest?" Then by all means try to talk them down off the ledge or make sure you're far enough away to not be in the splash zone.
  
  But this is someone saying "I want to make tools that non-experts can use to do productive things." I think it's not fair or reasonable to oppose that. Making computers more accessible and generally useful to the public is a good thing.

Yeah, OK. I just watched a complete novice ask ChatGPT how to go through installing node, pip, and creating a react app (of course ChatGPT being of 2021 suggested CRA). After 2 minutes, his confidence was soaring. And then he tried to run the react app and ran into an issue that required 2 days of troubleshooting for him to resolve. (When he asked me about it I told him he could have just deleted the file and moved on.)

So, yah, just let the CEO type in the code into the magic box, what could go wrong.

pip? What would that be used for
- Sometimes I serve apps with Flask, he wanted to get into some python as well

Get ready for a bunch of unskilled people making the shittiest apps imaginable.

We currently have skilled coders making the shittiest apps imaginable, due to shitty direction by management.

And this will make those apps look amazing in comparison.

I have no idea what this company is...who this person is...and by the looks of it, I'll never need to.

Replit is an in browser IDE. I met the founder once at a tech event. He struck me as a little arrogant.

Ahh, replit. Wanna know how to completely destroy your service, community, workforce and reputation in less than a year? Ask them.

I used to use replit at the height of its popularity. If any of you know what Glitch was, it was like that but way cooler.

You could program in virtually every programming language, write Discord bots, make games, talk to people about coding and get viral from your programs. It was basically programming but with social media and with extremely easy to use text editor and theme support basically made Replit the #1 place to program on the Internet.

Then, the AI boom happened. And replit wanted quick cash immediately. So they up their subscription prices slightly with Replit Agent, basically ChatGPT but for programming. They then removed their school system, and decided to stop virtually all community support. Now a lot ot the top projects were AI based but at least you could progran right? Not anymore.

Halfway through 2024, Replit decided to kill the platform in one fell swoop. They restricted all free users to only 3 projects, 1 GB of storage and severely throttled CPUs as well. This made the replit userbase tank. They removed all community features and essentially ghosted everyone. Anyone with a brain abandoned replit and just programmed on their own computers.

It's so sad what happened but hey, at least replit made people who would never program once in their lives programmers through ease of use and social factors.

And all they had to do was not be completely idiotic and greedy.

I wonder how well & fast that is going to age for the maintenance of those “app”

Ha! « CityTv » immortalized in that gif!

Sounds like they never cared in the first place

We don’t care about professional coders any more. You know, the people making all these “breakthroughs” we are so happy about.

Black hat hackers are drooling at the thought.

I have a novice programmer friend who created an app with authentication with AI. It used create-react-app (deprecated for years now) and had him roll his own OAuth layer manually with Express. When you refresh, you get logged out.

It's an awesome project to learn some skills, but only a seasoned programmer would know to never roll your own auth, or to not use CRA, etc.

We are in for some interesting times.

My mom is totally ready. She's in her presidential age. She can still talk pretty good, walks and is very independent. She has an iPhone if that helps.

How many active apps does she have at the same time. My parents typically have around 50 and then complain the phone is slow.

Every damn weekend for the last 5 years
- Arthritis, Is that an app? She complains about it all the time. But she could be a great AI programmer!

To explain the clickbaity title: it means their product won't target professional coders, not that they won't hire them.

Honestly, that makes sense, since anyone who knows a bit about Software development can see that handing off control of your app to a large language model in a way you have no clue of what's going on in the back is not sustainable at all.

Customers could, in theory, use Claude directly to create software, but then they’d have to handle everything else that goes along with it. “What you’d have to do is pay for Claude, go to AWS to start an EC2 machine, go into that, install Git and Python. Already, most people are just gone at this point,” he said.

so their competitive advantage is not having to start an ec2 instance lol

I hope that some of the coders losing their jobs to AI find a way to poisoning the well somehow and create tools to undermine the technology, the companies paying for it, and the oligarchs that want it.

What happens if someone (OpenAI) makes an LLM inject spyware into the code? Who would be able to read the code and figure that out if you have no coders?

You wouldn't even have to reach as far as malware. All software has bugs. To think that AI will produce perfect bugfree code because "it's a computer" is laughable. So inevitably there will be a need to debug the code, across servers, filesystems, databases, APIs, you name it. In tens if not hundreds of thousands of lines of code, which might even be compiled. Surprise, an LLM can't do that.
Nope!

Do I smell pump and dump?

Reads like a puff piece.

That's because the tech industry has a deep vested interest in this type of bullshit succeeding. So it's always presented as the most awesome thing ever. Did you see the not-a-sex-robot presented at CES this year as if it was basically human rather than a freaky uncanny valley Realdoll? Even the guy who they had do the video was clearly having trouble keeping a straight face while he praised it.

https://www.thegamer.com/i-cant-stop-laughing-at-the-175000-girlfriend-robot-from-ces-aria/
- So that's where all the new jobs are going to come from after the economy goes to shit because of climate change, oligarch hubris, etc.!
  
  They are going to pay the unemployed people pennies an hour to humiliate themselves by dressing as robots and doing all sorts of menial tasks for them. It's going to be like Sleeper meets Gone With The Wind.
  
  (I'm joking at this point, but this could make a good plot for a dark sci-fi comedy.)

Not like they're already out there.

Good luck with that, bud.

That site went from useful to a complete pile of trash. Maybe the chat bots can use it or something.