41% Increased Bugs With Copilot

Recorded live on twitch, GET IN
Article
shenisha.subst...
By: Shenisha
My Stream
/ theprimeagen
Best Way To Support Me
Become a backend engineer. Its my favorite site
boot.dev/?prom...
This is also the best way to support me is to support yourself becoming a better backend engineer.
MY MAIN YT CHANNEL: Has well edited engineering videos
/ theprimeagen
Discord
/ discord
Have something for me to read or react to?: / theprimeagen
Kinesis Advantage 360: bit.ly/Prime-K...
Get production ready SQLite with Turso: turso.tech/dee...

[@EduardoValdez-f4j]

The fact that AI can't help Junior developers should tip you off to the fact that even with AI, a normal person cannot write production level code, a massive blow to those thinking that software devs will get replaced.

[@RealTwiner]

One llm doesn't prove anything. Copilot sucks. It really does. I'm a software dev with a good amount of experience. Using claude, ive developed small applications without hand writing any code, just describing the problems clearly. Small problems like having it build ui controls that have been built a billion times it gets right the first time every time. The thing about that is, if you're doing a good job writing code you have a lot of little black boxes that don't have a ton of connections to other parts of the software. Obviously however, that doesn't describe all types of problems that needs solving, and as stuff gets more complicated, hands off auto generation ceases to be a thing. At that point it becomes more of a way to Jumpstart a problem. Where it REALLY fails however is architecture, which really requires systems thinking which LLM are not all that good at yet, and thats compounded by the fact that the context size of LLMs is limited, and for projects that hit real world needs you simply can't fit it all in the context. Even so, you can still speed up code generation if you're using the tools intelligently. But again, copilot sucks, it consistently gives weak and wrong answers. Ive only used those and local versions of llama, which I haven't reaaaaally put through its paces, so I donno how chatGpt really stacks up.

[@the_derpler]

@@RealTwiner I describe my problems clearly every day, with C :)

[@my_online_logs]

@@the_derpler maybe your way in describing it. i have no issue working using rust with claude that rust is more annoying than c and has less example and libraries out there compared to c

[@musashi542]

@@my_online_logs yeah nobody cares about your hobby project , were talking about the real world here .

[@OzzyTheGiant]

@@RealTwiner Even then, I don't think that will ever happen. There's no way that AI can reach that potential where it can understand context and architecture at such a high level as a human. We're talking about replicating the power of the human brain. We simply do not have the power to create that.

[@TheMiddleMan890]

I really need to know why Prime refuses to highlight the first and last character of a given paragraph. There has to be a reason, right?

[@Tymon0000]

He thinks that's cool.

[@nikarmotte]

He finds symmetry he says. I mean highlighting everything is also symmetrical. But I mostly listen more than watch, so I didn't even know until he mentioned it.

[@TheMiddleMan890]

@@nikarmotteCool! It doesn’t bother me or anything it’s just something I always notice

[@TomNook.]

Because theres a chance, depending on website formatting, you could select entire paragraphs/page/unrelated text/ads.

[@hakerananasek]

That way he won't accidentally highlight the entire text on the page

[@georgehelyar]

Using AI to write your code is basically pair programming with a junior developer, but without helping someone get better in their career.

[@kesky6363]

Well in a weirdly roundabout way you are helping it get better

[@ianjackson2696]

@@kesky6363not really. If you use the shitty code it gives you to write more shitty code, you just make it worse

It's more like pair programming with a really smart goldfish that pops hallucinogens.

[@ws9691]

💯

[@SoulSin]

Days passes and people start to realize that the whole "AI will substitute dev workers" was just another different excuse to fire people.

[@maaikevreugdemaker9210]

I use you in my playlist as watch later so that I can listen to your videos as a podcast, it's really great thanks for that I subbed.

[@DonaldSMay]

In my experience it really doesn't save time because any time savings are offset but subtle errors that it can generate that take time to debug. Until we can generate code that is trustable and correct then it is not worth the headache.

[@ThePlayerOfGames]

LLMs like copilot don't care about structure or facts, they care about the statistical correlation of text snippets. Copilot can't generate functional code, you need a different neutral network that also weighs the validity of outputs or understands logic logically in order to do what you ask Which is cool, but that's gonna take another half century

[@Excalibaard]

The most annoying part is that those co pilot ramblings take precedence over suggestions from an LSP that are based on actual type information. To be fair, I don't even expect my own code to be trustable and correct, though that could also be fully attributed to using TypeScript. I expect even less from a LLM that keeps imagining enum members and non-existent keys because 'it reads nice'.

[@SeeAndDreamify]

Amount of bugs seems like a very hard thing to measure. I belive that most codebases has tons and tons of bugs in them that the companies don't even know about. So you could get the same kind of effects as you get in crime statistics, where numbers sometimes go up not because there are actually more bugs, but because more of them are discovered/reported.

[@catfan5618]

That was also my first thought. How the hell do you measure that at all? Counting commit messages with “fix” in it?

[@schillville]

Also, how many times something is reported as a bug when it is actually something not accounted for? We've all been there...

[@panlis6243]

That, and there's not really an easy way to quantify "seriousness" of a bug. Like you could reference a wrong variable somewhere or you could introduce a memory leak in some very specific hard to notice edge case. Technically both of those could be "1 bug"

[@KillFrenzy96]

When it comes to AI and coding, I find that if you ever need to prompt the AI to write code for you, you're not doing it right. The time taken to understand, review, and bug fix an AI's work should not be underestimated. AI is more useful to enhance your own skills, not replace it. Therefore, I limit my AI use to locally hosted model for autocomplete on crack.

[@schillville]

My experience with coding assistants: - Wasted time reviewing generated stuff. Usually, it comes with errors/hallucinations. To be fair, I think this is expected since it is trying to "guess" what you need, but it breaks my concentration. So it's just faster to write by myself. - Most of the boilerplate it eliminates was already baked in IDEs for a long time, so it becomes kinda redundant. - Junior developers using code assistants is a mistake IMO. I've also seen some colleagues talking about their companies banning said tools for non-senior developers. So much time is wasted during code review because these developers don't even understand what the generated code does. And don't get me started on how many times I reviewed code that doesn't even work. I've used AI to help me learn some stuff outside of work. But I take everything with a grain of salt and always double-check the provided information. But as coding assistants? Eh, not really worth the hassle. Maybe I'm getting old lol.

[@MRM.98]

I don’t understand this logic. Should we ban google then too? At the end of the day people who are good engineers will be trying to understand what they are outputting…

[@Kalexand]

AI is great for senior dev who really know what he is doing. But for junior? No, a junior doesn't has the comprehension/experience to know when AI produce good code and when AI produce garbage

[@alexvisan7622]

Ai is great for CRUD app developers. It won't work for anything remotely complex.

[@Master120]

Yesterday i saw a comment on some video about AI, and the comment was a person saying that be was using AI to make a mobile app WITHOUT knowing programming or even the syntax 💀 If that's not bad enough, that person also said that AIs can replace programmers.

[@my_online_logs]

bruh im junior dev and succesfully using claude sonnet 3. to create fullstack audio visualizer website using sveltekit that can run without internet once the user visit it firstly that embedded to rust axum backend become 1 binary and created rust audio visualizer graphic and computation library that compiled to webassembly and import it to the sveltekit. the result is very amazing audio visualizer graphic that has very fast, smooth, and responsive animation. no lag hehe. i also created android apk, just a keyboard but not simple keyboard, but a full featured keyboard that has theme, also can ask to ai directly from the keyboard, translate using ai from the keyboard, ctrl key using kotlin. its the first time i code kotlin bruh

[@my_online_logs]

@@Master120 bruh im junior dev and succesfully using claude sonnet 3. to create fullstack audio visualizer website using sveltekit that can run without internet once the user visit it firstly that embedded to rust axum backend become 1 binary and created rust audio visualizer graphic and computation library that compiled to webassembly and import it to the sveltekit. the result is very amazing audio visualizer graphic that has very fast, smooth, and responsive animation. no lag hehe. i also created android apk, just a keyboard but not simple keyboard, but a full featured keyboard that has theme, also can ask to ai directly from the keyboard, translate using ai from the keyboard, ctrl key using kotlin. its the first time i code kotlin bruh

[@Haise-san]

​@@Master120yeah but... That person probably doesn't have a job as a software developer, nor they will ever have one

[@kuhluhOG]

15:20 Here's one big mentality difference between US-Americans and Germans: The former figures things out as they go. The latter figures things out first and then goes. That's a pretty fundamental cultural difference which you can even notice during vacations. But obviously this is a generalization, so you will always meet people who aren't like this.

[@triplea657aaa]

I find that it's really only useful when I know exactly what I want and it's just a more sophisticated tab-complete or if I'm so tired/migrained that I can't put together coherent code and something is better than nothing.

[@JohnDoe-bu3qp]

I don't know if Prime reads these, but often I don't like videos because my goddamn tv just closes them as soon as they are done, and then I can't like them anymore. So either I have to like videos before I've watched them completely, or I have to find the video again to like it it was a pain. This is why so many big channels have a song or something else that is always the same at the end of their videos.

[@ThePrimeTimeagen]

oh. that is wild

[@evancombs5159]

@@ThePrimeTimeagen yeah KZbin doesn't really consider the user experience differences between watching on a computer and watching on a TV. Having an outro makes a big difference for TV viewers.

[@rad1an_]

also in playlists (e.g. watch later). youtube just skips to the next video, not giving a chance to like/read comments

[@krellin]

its useful for shit like "generate compose yaml for this and that" and giving you quickstart nothing more... definitely not worth the hype, but can save time.

[@zhacks_admin]

..and for things like writing backend schemas/models, and data typings.

[@nicolaska1761]

You should try claude, it's miles aheqd of copilot in terms of both good practices and code trust. You can give it a lot of context as well

[@Theepicfrenchguy]

Front end wise, it has been helping me a lot for repetitive stuff. It definitely doesn't work well if you don't even understand what copilot is suggesting . Or if you use something very opiniated like tailwind.

[@suede__]

8:00 That comment on founders is spot on. I have a family member who was an architect and loved it, but realized to make any money he'd have to become a principal and/or start their own company and then they would end up stop doing what they loved.

[@kevinryan2992]

The single biggest benefit of AI for me has been to help with my RSI especially while re-learning to type on an ergo keyboard. Its great for filling in boilerplate for function signatures and stubbing things out and then turn it off when it comes to the actual implementation

[@fengziya]

This is me. For multiple times it has generated code that look fine at first glances but then I would waste hours debugging because of that exact piece of code, so I just turn it off by default now and have a keyboard shortcut to turn it back on temporarily to generate a small snippet I don’t want to write. It’s good for really simple stuff sometimes but don’t have it on at all times.

[@MartinWoad]

The question isn't if coding assistants will increase the productivity of current developers. The question is how much not having them is going to decrease the productivity of developers from the post-coding assistant era.

[@nikarmotte]

Why would there be such an era to begin with though? Genuinely wondering why you think that'll happen.

[@no_name4796]

I never used ai assistant tools. I simply prefer to ejaculate code by myself, then having to check whatever alluction ai comes up with. And it's easier and faster for me. For once, not using copilot, allows me to enter a flow state, where i know exactly what i am doing, and i am able to produce code (which mostly work, ignoring the stupid mistakes everyone does when coding lol) And also, waiting 1 second for text to appear it's just stupid lol. In 1 second, you could type 1 or 2 words at least

[@Testvvjnb-ci3zl]

I don't see any reason why the 'era' would end

[@williamdrum9899]

Aren't compilers just a "coding assistant" then?

[@MartinWoad]

@@nikarmotte Because juniors are now using ChatGPT left and right. And one day they will be seniors or at least supposed to fill their roles.

[@GIRcode]

Does it say whether it's a 41% increase in the amount of bugs or the likelihood to make bugs. If copilot increased code production significantly over 41% it might not be as bad an issue as the headline makes it out to be.

[@MatiasKiviniemi]

I founded a company and realized even though "I can Excel" I can't be the business guy. Mainly because I''m an introvert and cold calling people I once met to ask something took half a day of ampping up and rest the day something pleasant to calm down. A real business guy makes a dozen calls before I dial the first one. In fact that's how he calms down after something unpleasant.

[@tictac7877]

It’s already been said that it depends on the experience of the developer. I would say I am an experienced SQL developer and an intermediate typescript developer. When coding in typescript I would ask copilot to give suggestion of what I require and ask another more experienced typescript developer in our team to review to ensure it’s good and not blindly trust copilot. For SQL I would make some changes to copilot code if I think it would be better doing it a different way or to fact check it for performance because some SQL can cause performance issue. Our team use copilot and it seems to reduce code completion time with suggestion, therefore overall save some time.

[@JustinReeves]

I like reviewing code. I hate having my code reviewed.

[@ChrisIsOutside]

That's insane to think that it doesn't help and save time. I was a 10x developer and now I'm an 11x. But seriously, if you plan your business logic/architecture etc... Copilot should be used more like, I need to type X here but instead I am going to hit tab because I can see its writing what I want it to. done, saved 30 seconds. compound that over a 200 hour project and the saving is insane

[@VectorNodes]

I’m not really sure if 41% is even as high as it’s going to get. So many newbies are leaning on and schools are permitting the use of ai in education I worry too significant of a portion of people entering both the workforce and public facing code that we’ll see a major increase in bugs but also a surge in incompetence feeding extra trash back into the models

[@StephenSchlie]

I wonder if there is a disparity between junior devs and senior devs (I assume there are more junior folks out there) that the junior devs get more enjoyment because as you said it makes them feel more productive, but then ideally a senior dev reviews the code, sees it is 98% what they expect, merges, then has to help clean up the code when it breaks in subtle ways that the junior dev doesn't understand. This leads to more churn and toil work to fix dumb bugs that shouldn't have made it out.

[@Rohinthas]

As a german dev, my impression is that germans are a lot more skeptical when it comes to the promises of new technologies and a little bull-headed about doing things yourself. Sometimes that preserves and deepens meaningful craftsmanship, sometimes it leads to trailing behind the state of the art.

[@fastestdraw]

There are ways to collect good data through questionares, but they need a bunch of data to estimate how much people are lying to you, which is very hard in any field that hasn't already had a ton of study. I still prefer the 'analyse landfill content' methods to figure out how people are behaving rather than asking them. I'd love to have an analysis of abandoned Github projects, mapping copilot content put into them. Knowing if its rookie devs, or big teams getting frustrated, or whether the issues with GPT spam has tipped the scales on maintaining projects, and whatnot could be a really interesting analysis.

[@WorldFactions]

We have been doing interviews at my work all week for a few positions. It's amazing how many people now have years of experience being a software developer, but they cannot even code a simple fizzbuzz when asked. It's really depressing.

[@mattymattffs]

Programming outside of work 100% increases your burnout even if it's a project that you like

[@draakisback]

I mean this study is pretty obvious even though 800 people is not really a statistically significant amount. You would have to have many studies over a longer period of time to truly understand the impact of AI on programming and code quality. My hunch which is of course anecdotal is that it really depends on the domain and that for the most part AI is pretty good at helping out with boilerplate and other repetitive operations but it's not geat at helping you build something that is new/unique. People seem to forget that AI is limited by the data it's trained on as well as the quality of the prompts that it's being fed. A junior developer without as much experience is unlikely to understand the details that they might need to look out for and they likely cannot use the AI as efficiently as somebody who fully understands the problem that they're trying to solve. Also, while AI might be able to help with general development because of the sheer amount of data that it was trained on it probably can't help you with something much more niche.

[@gammalgris2497]

I find that I not only have need of code documentation but more and more have to look into 3rd Party code (I.e. library implementations) to avoid nasty surprises (i.e. error handling, runtime behaviour). I also need good examples of how to use libraries and frameworks but these are generally lacking if you find some. So it takes time to create code that you can use. That's one of the reasons I have still a hard time with LLMs.

[@FabulousFadz]

18:06 the person I trust the least is the one who is trying to sell me something. I'm just chilling then they say, _"dude, you need this!"_ nah... Should be missing me with that.

[@gracjanchudziak4755]

16:24 I'm one of this 3%, I have been learning coding for 6 years, I have 2 years of commercial experience, I have no drive to test this. I had experience with doing reviews of guys code which was generated by AI and it was distasteful, amount of bugs, poor designe, poor quality, there was used really weird librares like ast in Python for transforming dicts and quite frankly, I'm disgusted, they lied to me many times that it's not AI and I was forced to do reviews for AI everyday and act like nothing happened, baceuse my managment doesn't care. I don't want to see this shit never in my life.

[@avocadoarmadillo7031]

Hahaha. I have a few team mates sold on LLMs. We are working on a project to evaluate LLM efficiency at reading death certificates and coding out outputs for national drug overdose databases. Mostly accidents, some murders and ____ercides. Team puffed themselves up, the best performing model is 79% accurate at coding boolean and categorical variables. I created a baseline model that returns 0 for all booleans and categorical variables. 90% accuracy, problem solved : P

[@quickdudley]

Regarding how people use playlists instead of subscribing: there are actually a bunch of channels I really enjoy but I don't subscribe because the KZbin algorithm already pushes them enough and I don't want the stuff I've actually subscribed to to get drowned out even more than it is already.

[@Thiccolo]

AI chatbots may not be helping developers, but they are most definitely helping the rest of the IT industry. You had thousands of workers who had no idea how to code now all of a sudden they have something that can make scripts on a dime. If they work well enough it's fine.

[@PoppySeed84]

AI helped me build a usable C++ FLTK GUI for testing a Khet AI written in C. I just wanted to get something up that would allow me to set up the board however I want and modify params when its running. Built in hot reloading so I could compile and side load the C lib. AI absolutely helped me get it done in about half the time it would've normally taken. Just asking simple questions like how do I X in FLTK... then modifying its output to fit my needs. I haven't coded in C++ in years. That being said, AI still lied to me a lot. I've just had enough experience programming to know when it was going off the rails.

[@theJellyjoker]

The satisfaction comes for finding more bugs. Churn more code to find more bugs! More bugs! More bugs! More bugs! More bugs! Bugs! Bugs! Bugs! Bugs! Bugs! Bugs! Bugs! Bugs!

[@connormullin4547]

Using copilot to write code feels like explaining what code I want to write over slack to a person who has like 2 months of programming practice and trying to get them to write it for me instead. In the amount of time you spend giving paragraphs of clarifying details and correcting mistakes you could just write the code yourself many times over.

[@TesterAnimal1]

That’s insane. Do people not know what code they want to write? Do they just let copilot write anything and accept it? Sometimes it guesses the code I’m going to write and that’s good. Sometimes it doesn’t, and I WRITE THE DAMNED CODE.

[@DDanV]

I use llm's to create structured readme files and little else. It creates the markdown all neat and tiddy and I just need to give a prompt with useful info and interate as I see something missing or off.

[@mrpocock]

The increase in bug rate, if true, tells me 3 things -- devs using copilot are blindly trusting llm-generated code, they aren't writing (useful) tests, and the 3rd party review process is crap. So perhaps this effect is that increased code assistance is highlighting and magnifying existing skill issues and process flaws?

[@perzyk7720]

I recently got copilot education license and I must say that for algorithms it's good but for complex engineering it may stumble

[@anonanon6596]

I don't like this language "41%" Remamber this one contraception drug that gave one person in a million side-effects, and then there were headlines like "the new and improved version has a 100% higher rate of side effects" This language is missleading. What is the base line? What is the deffinition of a bug? etc...

[@squisherderheld]

If you get your new features done in 40% if the time it may be a fair tradeoff. Let others deal with the fallout

[@GuilhermeTeixeira]

Coding is fun. If you accept that Copilot takes it from you, that's because for you coding is somewhat painful. Coding is painful to those who select text using mouse, who need to click several buttons to achieve simple stuff etc. If you use Vim to code, you don't want the act of writing code to be taken from you. That's why I don't use Copilot or any other "AI" tool.

[@ryokuhasu9699]

I replaced an entire byte array buffer class in one of my personal projects with one written by chat GPT because it was error free and more efficient, but I wouldn't use it to write the entire project, that would be absurd.

[@hoyoreverse]

depends on where you use it. For example in C++ it's common to ditch std stuff, and implement your own String class and dynamic array (list). Saw this in a lot of C++ games I reversed.

[@jameslay6505]

I feel like lgtm reviews happen when you look over a PR and it literally looks good to you. There's nothing that is obviously wrong, nothing that is going to break the system, and the only things you could critique are nitpicky, pedantic, or bikeshedding issues. You just don't wanna steep to that level.

[@sIippo]

last time i was this early, prime was still working at netflix

[@Teodor-ValentinMaxim]

It doesn't improve code quality, but it does damn well autocomplete some variable names like crazy. It feels good sometimes that you don't have to start typing some variable names just to wait for the LSP suggestions to pop up. If it were free or 5$ instead of 10, then I would've continued my sub.

[@abc123evoturbobonker]

2:00 You try using a tape measure in the dark..

[@CrackSt0neTV]

Did he really just say 'rewrite tests'?

[@howuseehim]

Sometimes, gpt gives me shit code so confidently I don't question it

[@elcapitan6126]

6:58 lip-smacking intensifies

[@dmon1088]

I'm addicted to them and have wrote hundreds of thousands, if not millions, of LOC over the years "manually". To be fair, if I didn't have Cursor I wouldn't use any, Copilot and such won't cut it for me, I'm very picky when it comes to interaction system & design. Autocomplete is kind of meh, I just use it for refactorings cause in Cursor it jumps all over the place. That said, I would not allow a junior to use these in non-personal code, you need to understand both the LLM and the thing you are doing fairly well in order to not shoot yourself in the foot constantly.

[@drxyd]

It's a really good autocomplete. That's it.

[@BusinessWolf1]

I use it as autocomplete and don't tab more than like 2 lines at a time, but 99% of the time just 2-3 words and some special chars

[@mattymattffs]

I thought everyone was using co-pilot to do their PR reviews so of course looks good to me is going to go up

[@canofpulp]

Microsoft recall is now enabled for all users by default in the new windows 11 update!!

[@ws9691]

My AI girlfriend left me for Claude

[@Nikolapestanac]

I've never used copilot, but when i heard "you'll type an if ....and wait", my jaw dropped; i never expected a tool to be able to affect a programmer's habits so deeply.

[@quickdudley]

If I'm using a slow enough computer (or even an OK one that I'm running too much stuff on at the same time) my workflow is "type some stuff... wait for it to actually appear on the screen"

[@SimGunther]

Why would this tech be theScapegoat for the board's bad and shortsighted financial decisions? Their party can't last forever....

[@lx2222x]

Copilot is like a fast car, not all people can drive.

[@KvapuJanjalia]

I've been reviewing sh*tty code written by people for years. Nowadays I review sh*tty code written by AI.

[@JasonParmenter]

It's fine for small stuff. I want to create a GUI wrapper for modules on python, ima just use AI tbh

[@justanothercomment416]

I've been telling people this since the technology came out. It's obvious if you've ever spent five minutes looking at the generated output. Top tier developers don't use this nonsense. For good reason.

[@eltreum1]

AI is not going to write good software easily. LMs are an overpriced probability engine that only knows what you teach them. What if it was taught coding from bad examples? I think it's more useful for things like 'Give an example of how I do operation X in C.' or 'Here is a function in C. What would be the equivalent code, if possible, in Go?' or 'Examine this function and suggest a more efficient way to do the same task if it exists.' The problem is LMs have no context of your app and what it needs without you writing a user story in a digestible AI prompt and hope it understands.

[@superdooper6969]

i spent whole day today chasing bugs that auto complete put there and totally slipped by my eaglevision

[@veekthorcodes]

i am literally facing this dilemma rn, I feel like gpt is not actually helping me as much as I think...

[@spiritkid9109]

STOP LISTENING TO THE PEOPLE WHO ARE SELLING AI FOR THE NEWS ABOUT HOW GOOD AI IS AND HOW WELL DEVELOPMENT IS GOING

[@jaymason7097]

That's copilot. Imagine the stats for Cursor.

[@storymode9085]

I use AI to understand, not to write. Even when I find out "how" to do something, I have to modify it to my needs. The concept and goal is what I take from it, but even as a 8 month old programmer, i have noticed that sometimes what it is doing does not fit my use case. I end up asking, what is "concept" see examples, then ask it for more granular explanation, then figure out how to go about it in my code. Copy/pasta??? Get Reked

[@evancombs5159]

This is how all programmers learned to program, just before AI we used Stackoverflow. I have found AI to be a much better user experience to Stackoverflow that helps me to get answers much faster. AI for code completion is still very iffy, but AI for getting questions answered is better than the alternatives.

[@GuilhermeTeixeira]

15:04 it's because Brazilian devs are more talented Also BRAZIL MENTIONED

[@Oler-yx7xj]

Programmers finally trying to do scientific studies

[@rumplstiltztinkerstein]

It would be so funny if Github Chief Operating officer said that Github Copilot was mid

[@jasoneisele5920]

Huh, who could've thought

[@myronm6507]

WHO WOULDA THUNK IT! :D

[@jameslay6505]

Feels like this article was written by AI.

[@FandangoJepZ]

Review fatigue with shitty AI generated code is real. Can’t even be bothered pointing out issues because the writer won’t learn from them. They’ll just generate the same problems again

[@JakobKenda]

most CTOs are developers, no? or is my experience that out of the ordinary?

[@progste]

wait you mean that random code generation leads to bugs???

[@FeLiNe418]

If I uses copilot, I'd spend a good chunk of my time reviewing the code produced by copilot which is a waste of time.

[@karlssberg]

For me, ai is just a better version of stackoverflow - it helps with my skills issues

[@KhaanKuinton]

It’s almost as if you need to actually know the #%^ you are doing when you are programming and/or what needs to be done.

[@Mammux]

I liked and subscribed.

[@krux02]

This study is no surprise to me. If you want to increase productivity, you need to train the neurons in the brain of the developer. Get them a better understanding of the codebase. Train them to know every piece of it. Generating more code with less brain activity doesn't help at all. If AI starts some day with code simplifications. Automatic refactorings to delete duplications, I am on board again. For now I am out for AI in programming.

[@tylerwatt5651]

OMFG. AI in Xcode is cancer.

[@Zonno5]

instead of writing and designing code you are now doing 100% code reviews

[@bravo90_]

could be from the higher output of code

[@meryplays8952]

In my case marginal benefits.

[@my_online_logs]

copilot is based on gpt, so its suck. claude sonnet 3.5 is different story, always complete the task using rust

[@RayCh]

80% of work force of US might be Indian Immigrants, Just kidding, so the similarity.

[@creativeb549]

:( I have a crush on you, stop being smart.

[@AndrewBrownK]

Have to cut out 25% through because of the incessant lip smacking

[@nickwoodward819]

is copilot representative though? last i looked it had fallen pretty far behind

[@axb8886]

10 mins ~20 likes? Lol. FellOffAGean

[@bonsairobo]

c++ compiler errors are *not* good

[@bearwolffish]

15:20-15:26 this ain't a good look no matter what way you spin it. Had to cap that timestamp cus there is some racial joke after and I ain't tryna get dragged in.

[@asi_karel]

it went from 2 paragraphs with 1 errors to 2000 lines in 20 files with zero errors in last year in the same time

[@llothar68]

autocomplete don't help a lot