> I've used Claude for many months now. Since February I see a stark decline in the work I do with it.
I find myself repeating the following pattern: I use an AI model to assist me with work, and after some time, I notice the quality doesn't justify the time investment. I decide to try a similar task with another provider. I try a few more tests, then decide to switch over for full time work, and it feels like it's awesome and doing a good job. A few months later, it feels like the model got worse.
I wonder about this. I see two obvious possibilities (if we ignore bias):
1. The models are purposefully nerfed, before the release of the next model, similar to how Apple allegedly nerfed their older phones when the next model was out.
2. You are relying more and more on the models and are using your talent less and less. What you are observing is the ratio of your vs. the model’s work leaning more and more to the model’s. When a new model is released, it produces better quality code then before, so the work improves with it, but your talent keeps deteriorating at a constant rate.
I definitely find your last point is true for me. The more work I am doing with AI the more I am expecting it to do, similar to how you can expect more over time from a junior you are delegating to and training. However the model isn't learning or improving the same way, so your trust is quickly broken.
As you note, the developer's input is still driving the model quite a bit so if the developer is contributing less and less as they trust more, the results would get worse.
> However the model isn't learning or improving the same way, so your trust is quickly broken.
One other failure mode that I've seen in my own work while I've been learning: the things that you put into AGENTS.md/CLAUDE.md/local "memories" can improve performance or degrade performance, depending on the instructions. And unless you're actively quantitatively reviewing and considering when performance is improving or degrading, you probably won't pick up that two sentences that you added to CLAUDE.md two weeks ago are why things seem to have suddenly gotten worse.
> similar to how you can expect more over time from a junior you are delegating to and training
That's the really interesting bit. Both Claude and Codex have learned some of my preferences by me explicitly saying things like "Do not use emojis to indicate task completion in our plan files, stick to ASCII text only". But when you accidentally "teach" them something that has a negative impact on performance, they're not very likely to push back, unlike a junior engineer who will either ignore your dumb instruction or hopefully bring it up.
> As you note, the developer's input is still driving the model quite a bit so if the developer is contributing less and less as they trust more, the results would get worse.
That is definitely a thing too. There have been a few times that I have "let my guard down" so to speak and haven't deeply considered the implications of every commit. Usually this hasn't been a big deal, but there have been a few really ugly architectural decisions that have made it through the gate and had to get cleaned up later. It's largely complacency, like you point out, as well as burnout trying to keep up with reviewing and really contemplating/grokking the large volume of code output that's possible with these tools.
Your version of the last point is a bit softer I think — parent was putting it down to “loss of talent” but yours captures the gaps vs natural human interaction patterns which seems more likely, especially on such short timescales.
I confusingly say both. First I say that the ratio of work coming from the model is increasing, and when I am clarifying I say “your talent keeps deteriorating”. You correctly point out these are distinct, and maybe this distinction is important, although I personally don‘t think so. The resulting code would be the same either way.
Personally I can see the case for both interpretation to be true at the same time, and maybe that is precisely why I confused them so eagerly in my initial post.
I don’t think the providers intentionally nerf the models to make the new one look better. It’s a matter of them being stingy with infrastructure, either by choice to increase profit and/or sheer lack of resources to keep n+1 models deployed in parallel without deprecating older ones when a new one is released.
I’d prefer providers to simply deprecate stuff faster, but then that would break other people’s existing workflows.
Point 2 is so true, I definitely find myself spending more time reading code vs writing it. LLMs can teach you a lot, but it's never the same as actually sitting down and doing it yourself.
I think it might have to do with how models work, and fundamental limits with them (yes, they're stochastic parrots, yes they confabulate).
Newer (past two years?) models have improved "in detail" - or as pragmatic tools - but they still don't deserve the anthropomorphism we subject them to because they appear to communicate like us (and therefore appear to think and reason, like us).
But the "holes" are painted over in contemporary models - via training, system prompts and various clever (useful!) techniques.
But I think this leads us to have great difficulty spotting the weak spots in a new, or slightly different model - but as we get to know each particular tool - each model - we get better at spotting the holes on that model.
Maybe it's poorly chosen variable names. A tendency to write plausible looking, plausibly named, e2e tests that turns out to not quite test what they appear to test at first glance. Maybe there's missing locking of resources, use of transactions, in sequencial code that appear sound - but end up storing invalid data when one or several steps fail...
In happy cases current LLMs function like well-intentioned junior coders enthusiasticly delivering features and fixing bugs.
But in the other cases, they are like patholically lying sociopaths telling you anything you want to hear, just so you keep paying them money.
When you catch them lying, it feels a bit like a betrayal. But the parrot is just tapping the bell, so you'll keep feeding it peanuts.
Freakanomics podcast had a recent episode regarding Cheating with PEDS, and interviewed the (former) head of the Enhanced Games. At one point, he discussed the benefit for society because athletes would be monitored for 5-years post performance.
To me, it seemed like a modern day tech-take of human cock-fighting.
In my opinion, the problem with PEDS isn't adults taking them if they would just admit to taking them.
The problem is with adolescents taking them. Adolescent boys see a really nice immediate payoff for taking PEDS (better musculature and better sports performance->more popular) while the downsides are in the future. It's really hard to fight that.
Even when I was in high school several decades ago, we had a handful of people on PEDS. And we were a tiny school with no significant sports programs. I can't imagine what it's like now with social media pushing everything.
> In my opinion, the problem with PEDS isn't adults taking them if they would just admit to taking them.
The incentive to cheat and hide was one of the points from the podcast. In Cycling, in order to win, you have to compete with other cyclists who are doping, and doing so in such a way that they are unlikely to get caught. In order to win, you have to dope and not get caught. Youre not forced to dope, but the option is there, and yours to take should you choose.
Honestly PEDS are stigmatized and under-researched for the performance enhancing aspect. They have undoubtable side effects - but how much, why, etc. is kind of meh from what I saw when I was looking into this, bro science is best you can get. Few studies here and there giving people modes test boosts and measuring athletic performance.
Not saying we should be promoting them, but if we can eventually get to the point where we eliminate the really bad side effects and get most of the benefits it's going to be a great thing for everyone, the next thing after GLP-1.
I do not have the background that allows me to make medical decisions based reading published medical articles, so I have to trust my doctors advice, and seek 2nd opinions if I'm not convinced.
My issue was the disingenuous use of a "5-year post compete" monitoring as justification for Enhanced Games.
I’m on a field trip chaperoning my kid. I get a couple slack messages asking for some tweaks to a UI. I type a couple words into a Github AI Agent Session while riding the bus. Fixes are deployed to our staging env in 10 minutes.
> I'm definitely a bit sus'd to run OpenClaw specifically - giving my private data/keys to 400K lines of vibe coded monster that is being actively attacked at scale is not very appealing at all.
Ignore turning lose agents on the internet that are capable of pulling in unchecked data into it's context window.
We're also seeing that the people up top are using this to cull the herd.
reply