truncate

That’s a norm at my place as well. But it’s not free. We have definitely given up on quality (both code and general stability) and fully understanding what’s going on. People are generally working more as well, because of all the layoffs pressure.

Industry is betting that we don’t need to. I still consider it a bet because Claude Code this scale has become norm just for a year.

>> what people are sending to Claude that is so secret

Its the same point used against privacy. What's so secret you are doing that you need privacy. I think in the end, its about privacy and not trusting these model companies with your data. Facebook manipulated people behaviors with all the data they had, no reason AI companies wont someday decide to do that same, and they have far more intimate knowledge.

When it comes to coding, I also don't like the idea of them taking my money and potentially at same time potentially using as dataset generator.

I've mixed opinions on that. For example, there are products like Google Shopping or Google Travel etc. They show up bunch of stores where you can buy a flight ticket or a particular product from.

However, in the end most people just end up buying it from top 2-3 marketplaces anyway, not because its cheaper, but also because they can trust it, it arrives in 2 days, and you can easily return it if you don't want. So if I'm going to most likely buy something from Amazon anyway, why not just use Amazon web page. They've had one click buy button for ages.

Reminds me of Turing test. https://en.wikipedia.org/wiki/Turing_test

>> but I kind of wonder why we write articles like this

I think its a way to speak out and get it out there, than stay unspoken. Author as well others who share same opinion get some validation that they are not only ones.

I'm actively uninspired to write high quality code when using Anthropic/OpenAI models given the high chance I'm a customer as well as used as dataset generation tool for them.

But currently cloud does beat costs of hardware ownership, particularly with ridiculously high RAM/GPU/SSD costs....again due to these same companies.

I'm sure they do, but its not 0 or 1 thing.

Or, they can figure out something else out? I recall couple years ago when China didn't have enough GPUs (still don't?), DeepSeek team figured out how to train with less computing. IIRC they made Mixture of Experts mainstream and made really optimized kernels and clever use of PTX instruction set.

Seems like parent thread spoke from perspective of developer/tech savvy person.

But even from end-user perspective, maybe its hard to see change when you compare one year or so, but over somewhat longer period I definitely see improvements. Over say a decade or so, wifi/bluetooth/trackpad support has significantly improved. Over last 5 years, I think gaming has become huge and you can actually use Linux as gaming machine. In last 3 years or so, I think I've seen my 5k display using Thunderbolt work much better. Fractional scaling also works much nicer.

NVIDIA GPUs are not shitty to use on Linux, and AMD has been working great for slightly longer. UEFI/TPM works much better now as well, where I don't have to worry about booting in legacy mode when using GPU drivers.

I didn't have to think about any of these above. If I think harder, I can likely come up with many more.

As great LLMs are, they are no where close to any biological brain. We are not even close to replicating human brain or even brain of an animal. Let’s not add more fuel into this hype.

I think its just matter of time. We can tell, not because we are into tech or not-"average person", but because we have seen outputs of these systems enough times as well as actively use it.

AI is in growth phase, and as more people use it more they will be able to tell particularly as they interact first hand with it.

Going back to your original point, people don't hate "slop" because they think it will cost their job (maybe some people). People hate slop, because they are forced to read it most of the day and if you want to do your job well closely babysit, verify and rearrange. The overall expectation of quality has declined over quantity.

I'm not saying its not productivity booster. But the way its used to increase quantity and then map it to productivity. This results in to two things (1) more work for people so worse WLB (2) increased dissatisfaction with work because the quantity is high and maybe you can call it X% productivity boost, but maybe with X/2% productivity boost you may get Y% quality boost, wlb and satisfaction with your work.

How do you define average person here? I have friends who are not into tech, but use instagram or ChatGPT. They can tell. It’s just matter of using it enough times.

For the work, I don’t think it’s just “costing the job” fear. It’s so many more things. Someone everyone is even more overworked, the amount of babysitting they require which everyone just seem to not talk that much about. Design docs and code that are increasingly weirder to read, sometimes make sense and other time look like they make sense but not really, the docslop, commit messages slop, I can go on.

It’s mix of people issue where it’s race to bottom to increase utilization, which leads to less than ideal experience. Technology is great but it’s definitely getting more trust than it should when we accept the output. I think everyone would be much happier, and technology would have much better reputation if everyone just slows the f down and uses it bit more reasonably. What is reasonable, you figure that out by slowing down and actually thinking about the problem you are trying to solve … not just prompt your way out for everything.

I don't think they are mutually exclusive, that if you take one you give another. TBH, its more about the team you work with than startup vs big-tech company thing.

I don't know how many of big-tech have flat hierarchy, at-least not the ones I've worked at and the people/team culture have definitely better. Were there times there was bureaucracy and all, sure, but I personally would take take good WLB most of the year, over fucking up my health/relationships over making some random software on claude code all night that these days anyway most want to be AI slop. If I've the energy to work extra, I'd work on my own hobby software projects.

That said, this or that, its all personal preference. So no wrong way of life here. Whatever, you want your life to be.

TBH, having worked at all big-tech, startup and mid-sized companies, pretty much most has some major pros and cons of their own, and these days ever more so. With certain big-tech at-least there some chance of getting decent WLB.

A single agent can produce enough code that even a team human can't reasonably keep up with if properly reviewed. An army of agents, definitely not. Most of us this everyday at work. Some are ok with it some are not.

The way I see it, the people understand that software, somewhat similar to how most understand their cars - at a very high level -- there is wheel, steering, axles, engine, gas tank, you put petrol to run it etc etc. If it breaks down in middle of road, even with all the tools most wont likely figure out what's happening or fix the car, and eventually goto the mechanic.

I mainly use Apple one on MacBook, and haven't noticed any CPU or issues with that. I've another one which has more power, can't say I use it that much so no idea about CPU, but I think because of higher power draw may affect battery slightly (prob bigger issue on phones).

I like a good old headphone jack, but then the USB DACs these days are also super small and portable that I just keep connected to my wired sets. I have couple nice IEMs, and because my phone doesn't have jack, I've to carry around USB DAC anyway.

> But it doesn't give value, it prevents value loss.

It keeps thing operational. Software changes because requirements changes, the context it is used it changes, or just the iterative nature of it where features are rolled out over time so that users can immediately start getting some functionality if not all that was originally planned (MVP).

Maintenance work is neither virtuous nor waste. Its just nature of the things we build.

I agree however with the part on more talk about low-maintenance patterns and products. IMO its often the trade-off between velocity vs quality/technical debt. So it happens, industry is favoring more and more towards velocity for delivering features that often add little value to the users, just because $$$, competition and maybe the grind culture.

This is also one of things that needs to be well optimized to be any useful. Army of agents may get an implementation, but not sure it will be actually useful, and I'm also not sure if optimizing that implementation is going to be easier than more well paced use of LLM where person actually understands what code is being generated.

Depending what you are making; coding slows you down and think as you build stuff step by step. Its kind of like speaking vs writing. Generating code through LLM for me is like talking -- I love talking, but when I write something down, I'm more structured, think more about things etc etc.

People say, if you you write elaborate well spec'ed prompts, it will do the right thing. Coding in a way is doing that.

I guess I get what you are saying. Creating the first of its kind is "deep tech".

I really don't care about the definition or semantics here personally. I think creating first one is great, but then depending what it is creating 10th generation can also be quite a breakthrough. First IC had like 10 transistors, modern ones do billions and getting there probably require way more work, ideas and research than the original IC did. Things just don't improve in silos.

I don't think I did reiterate your point. You were saying unlike ICs, databases -- "But it's not creating something new that has never existed before".

The point I was making was like newer/better ICs pave way for new electronics/use-cases; so do better databases/storage software pave way for more new applications that wouldn't have existed before.

>> Based on their track record this lawsuit is not unexpected at all.

Their track record, and the narcissistic tone where they talk so confidently about how they are morally superior companies than others, and all they care about is public good.

>> It's not inventing the transistor or the integrated circuit.

Integrated circuit on its own isn't doing anything unless used for something. Better chips enable more use-cases, more electronics etc etc. Databases in similar way are enabling more applications for users. None of the apps or services we reliably use everyday at the scale (e.g. Google Search, Online shopping, Messaging) would not exist without solid storage software.

Not that I don't believe its possible to fix a lot of bugs, I also wonder what the actual dynamic was. Were the people in team working much more than usual as well? Given its Google, I wouldn't be surprised if there was an "internal push" to fix more bugs over next X sprints so that they can publish this blog and some manager can show impact and AI adaption to his superior.

Crack down on distillation, just for Chinese companies or is it ok for Chinese/US companies to distil? I find it hard to take Anthropic/OpenAI on distillation, because the way see it they started with "distillation" of another kind. They used all the content out there without consent of the creators and its still happening. Model distillation is just a different layer of abstraction, but same thing more or less.

>> It's annoying that we use broad terms to describe a set of technologies that in some ways can be problematic and in another ways are very beneficial. We gotta evaluate each of these as they come rather than talk about blanket bans.

I totally get it. I think few years, if some company said they record and transcribe every meeting/interview they take, it would be concerning. Now, its somewhat a norm for people to use these AI meeting tools which record everything you say and then go back to recording and exactly what people said. I'd call it surveillance than AI

I sometimes think its somewhat like opioid. Has its use, reduces pain, but then after extended use you also likely get addicted to it and starting having issues.

I wish README is clear whether they are using AI or not, and if so what the guidelines are. Not that anything wrong with using it, but given anyone with $$$ for tokens can do it, its nice to know what their process is etc etc. Gives me more confidence that its worth checking out.

Overconfidence. These people think they are much smarter than others to be caught.