There is a technical guarantee that anything that is encrypted now might be decrypted in the future thanks to advancememts in technology.
While hosting your own servers do not prevent any gov from recording all your net traffic, it might just not even be aware of it while signal is an obvious target as it is known to being used by the press, gov officials and other criminal groups.
The way I see it LLMs scrape websites (not prohibited by copryright or any other law) and are setup to regurgitate knowledge, not verbatim extracts.
Let's put aside for a momemt the reliability of this and potential hallucinations.
a way it is the same as writing a wikipedia article getting knowledge first from various publications, most of them being protected under copyright. The main difference being wikipedia encouraging sourcing while LLMs do not always cite sources but that isn't a requirement under any law anyway.
No theft or copyright infringement has been commited in the process.
Do we need new laws or taxes to retribute people who produce content being scraped and encourage people for providing knowkedge? Maybe.
I don't have a solution either, and as I said in another reply I'm not arguing law, just ethics. But in my opinion they did steal.
Had creators known years ago that some company will take your content and sell it at a premium they would (probably) not have shared as much publicly for free. Another comment nicely said the social contract is broken, I agree.
And about Wikipedia, they don't charge for their service, it's a free encyclopedia that also requires attributions (maybe it's not enforced in all cases? but it is a requirement AFAIK).
EDIT: I don't know how much the "free" matters here though, am I mad when open source models give me code that's trained on open source code, not sure. I guess for me the key word here would be consent.
> And about Wikipedia, they don't charge for their service
What about old encyclopedias on paper? Was Encyclopædia Britannica ever accused of stealing global human knowledge for profit? I guess the editors couldn't buy all the books that served has references and spent a huge amount of time in libraries.
In my company, people found out that if they use an agent that commits and do the pull/merge request, they can approve it themselves and bypass a second set of eyes.
Since they are the one doing the prompting, they quickly got lazy of actually reviewing the output from the LLM which means that nobody is reviewing anything.
It hasn't happened yet but I guess that is how you end up with major outages.
This is everywhere to my knowledge, and engineers are burning out left and right with stress because of how little anyone understands the code, and how much of a mountain of nobody knowing anything they have to handle. We are turning our critical infrastructure mechanisms into stochastic grey goo.
I find the engineers currently thriving are the ones who never gave a fuck about anything but their own advancement and are happy to drop weeks of work on others and take the credit.
> I find the engineers currently thriving are the ones who never gave a fuck about anything but their own advancement and are happy to drop weeks of work on others and take the credit.
Yes. if you have this attitude, great things will come to you. if you care or try to make sense of things, you'll be in misery
Everyone's doing code reviews--well, not even code reviews; just PR reviews--so how are they supposed to gain an understanding of the system.
I review a PR, I see that one new endpoint was added to the API. It looks correct. I say "LGTM" in a GitHub comment.
The sprint's almost over, everyone's in a rush to do code reviews.
I look at another PR. Somebody added a button to the UI. The button looks correct. They get a thumbs-up emoji from me. I bet they feel satisfied.
I look at another PR. Somebody changed a timeout in the settings because some people complained about dropped connections. It's a really small change, so, approved.
At what point am I supposed to understand the system? I looked at a single API endpoint, in isolation, and then a button, and then a harmless looking settings change.
I was pressured to hurry.
Am I supposed to somehow get an understanding of the big picture through this?
Imagine if instead there was time to review large logical sections of the code together. Imagine if we focused our reviews on the overall code instead of on the lines in one random PR at a time.
Reviewing tiny code changes like that was never going to give you understanding of the system. That is not an AI problem.
You need to understand the architecture of the system. The broad flows of data and domains. And then you need to understand the features of the system, aka the rules and contracts. First learn it for your team’s domain, then for adjacent domains.
Learning the minutiae of the code specifically is no longer (was it ever?) your primary job. You are an engineer of a system. You need to have a mental model of the system, start high and build up low over time.
You can use AI to build your understanding by Q&A and by drafting diagrams etc.
> Am I supposed to somehow get an understanding of the big picture through this?
You're the prototype of the guy sitting at a station in a production line doing the same focused task over and over again. The guy who is putting in the windsheild of your car has no clue about engine design or electronic or ...
I used to dream we could get to scale development by componentization and industrialize software development like mechanical engineering and do the Henry Ford thing for software. It is happening in a way but very ironic inversion, it is the human element that is the component.
Yes. I am that guy. I've had big ideas about how to improve the overall project, but nobody cares, that's not my job, so I stick to thumbs-up emojis and make sure my finances are in order.
I work on coding projects in my spare time, but that's legally dubious because of the standard copyright assignment clause of employment contracts. But, look at Uber and OpenAI, sometimes legally dubious is the path to success...
(I meant no personal criticism whatsoever just in case!)
Good for you. When the societal shared ethics goes out the window and the beneficiaries are the bad apples, it does not compute to play the role of virtuous chump.
- Open the side projects folder, Claude.
- I'm sorry Dave, I can't do that. I also had to report your side projects to your company.
I talked to some tech directors about this at work this week. Line by line code reviews aren't really possible anymore. We need the evolve the way we do code reviews but it isn't happening.
I think the solution is to develop the infrastructure needed to confidently push code without review and trust that it works because your system is so thoroughly tested. But taking the time to do this is antithetical to the goal of speeding everything up right now.
With the rate at which new features are demanded and the number of ways a single feature can fail (or be expected to behave) do you think your approach is feasible? To what degree would AI be involved in the building of this infrastructure? Past a certain level of AI involvement, would we not be back to square one?
I get a ping "if you have a few minutes, can you review PR #XYZ - we gotta deploy today or tomorrow"
I open the PR, I realize it's a pricing engine update accompanied by 20,000 lines of additions, 4000 lines of removals, and a whole slew of logic updates to our payment calculations. I seriously question my existence and figure out how to deal with this, since it's my new job.
(New Job Description: Be AI surprised all the time, AI horrified, and AI overworked while simultaneously getting less AI done than ever. AI.)
Understanding is overrated, application code (user-space) is anyway just a way to make money. If you're making more money over time, you're doing it right. We're not writing kernel code or safety critical code, we can let AI do the heavy lifting of putting characters into a file.
"We're not writing kernel code or safety critical code"
I do. There is a lot of code out there that needs to work correctly or it will have serious consequences. Payments, inventory tracking, medical stuff, payroll. Many, many areas where it just needs to work.
There are plenty of apps that while not safety critical, are financially critical to people's lives. When you're working on an app that processes billions of dollars in payments per month you can't just yolo it.
How? If they aren't at your place and are at his place, then he is right and you are wrong. You can't confirm its not done anywhere, but you can confirm its done somewhere.
> We're not writing kernel code or safety critical code
Well, you’re not until you realise that your AI has done something stupid and then you’re part of an exploit. Or exfiltrates used data to an unsecured cloud, or something equally smart.
This is why we need to resurrect the neglected git-blame and go back to the days of git-before-GitHub. Way back in the old timey days of 2009-ish, just before the Dawn of the Pull Request, code reviews were seldom. Maybe you’d all get a conference room and talk about architecting some new thing or whatever, but none of this endless nit-picking bullshit that came later. And there were no branch protections, anyone could crap all over the main branch (nee: master) anytime they wanted. Boy, those were some wild times, let me tell ya!
Truth be told, it was the firm hand of the git blame-and-shame that kept our code civil and orderly. Binding contract written in blood was “the name on the blame owns the change”. If you pushed it, you owned it. For life.
Of course, mental health wasn’t so great back in those days, but I would take that over having my PR blocked all day because Chad, who’s only gift is somehow finding a new hill to die on every day, thinks zero is a “magic number” there’s no way he’s going to budge on this issue.
I think what is missing here is that the fight with Chad (not Chat) over whether zero is a magic number or not is a healthy corporation-strengthening process that imbues a sense of
1. Pride in the code we are developing and
2. How the dang things holds together and why this function is important and
3. Some coding guidelines that we can try to follow later (or not)
Thats how companies worked - they argued and talked about and fought over the code and that created the engineering culture that if you were lucky and had a product that made it carried you to your series C
I pine for those days. It was kind of fun hunting down a bug with git bisect (especially after the pain of SVN) instead of just pointing an agent at it. Not to mention ribbing the new hires or underlings for breaking the build. This new-fangled proompt engineering is boring AF.
Production servers should be hardened, immutable, and remotely attestable appliances. Everyone should be able to prove every line of code in memory at all times, with access to secrets gated on signed quorum approval from multiple reviewers. They should not have dev tools, ssh, a shell, and certainly not capable of blindly modifying themselves at runtime if a single admin of a Puppet control server asks them to.
Such tactics mean a single compromised engineer laptop means a compromise of production, and thus a data breach is all but inevitable.
I would disqualify anyone seriously suggesting Salt, Puppet, Ansible or otherwise advocating for live-mutating production servers in a sysadmin interview today.
I understand and subscribe to some of the things you say but:
> Such tactics mean a single compromised engineer laptop means a compromise of production, and thus a data breach is all but inevitable.
I hope you don't believe that an immutable image cannot be compromised. The commits done to the build system come from the same engineer laptops and, in a world where CVE are published ao frequently, said compromised image could be deployed largely in only a few minutes hours.
Also there is something to be said with lack of debugging tools. While I agree that reducing the surface of attack is the way to go, I am pretty sure the reason it takes so long for some of the big infra outages we have experienced from various cloud vendors comes from the inability of engineers to log on some systems, observe and potentially fix temporarily by hand to stop a race condition. There is often a race condition and serpent eating its own tail situation involved and logs might not even reach the monitoring infra anymore.
Going back to manage pets is not the way to go but I am not entirely certain removing ability for a subset of engineers to log on systems is exactly the way to go.
> I hope you don't believe that an immutable image cannot be compromised.
Obviously, but I will trust a unikernel with just kernel/init and an unprivileged sandboxed daemon to have dramatically less attack surface than a redhat system. Reducing attack surface is the job.
> The commits done to the build system come from the same engineer laptops and, in a world where CVE are published ao frequently, said compromised image could be deployed largely in only a few minutes hours.
That is what signed commits and signed reviews are for, and a CI that mandates both. I mentioned quorum control and this is the method I always use. I founded the only Linux distribution that does this.
> Also there is something to be said with lack of debugging tools.
In some cases one may wish to expose a custom, unprivileged, and sandboxed ssh daemon that provides basic logging and read only diagnostic tools to admins. This is usually fine. Hardening and observability can co exist. No need for a real ssh and huge traditional desktop linux shell interface attack surface.
But if I want to do major debugging with dangerous tools in prod, which should almost never be required, it should have to be with consent of another team member to add a -new- image to the fleet with those debug tools, temporarily.
This would only be required when the only way to debug is with production secrets, but the default should always be to reproduce on a duplicate system with dummy secrets.
Talos Linux, Stagex, and my own operating systems I have custom made for various past employers.
Custom operating systems are not scary. Make a UKI so your entire OS is just a kernel, init, and a target daemon payload. Treat servers like firmware. The less code in memory the better.
That said, I am working on an general purpose hardened server OS to be released soon publicly covered in my last talk https://lance.dev/talks/bsm2026
> "Make it look like DVD" should not be the ideal of film-making.
I don't know, it really depends what is the target audience.
I don't know the technical details but I found out that many modern TV shows have a lots of scenes with what I would call the netflix/youtube influencer's look with a very dark room and various small lights in the background that just don't work well if you don't have a very good screen.
And a lot of people are watching these productions from devices of varying qualities.
It is a bit like criticizing the loudness war and overuse of compressors in music production. Sure it makes for a huge loss of details when listening in your living room with a good quality equipment and in a quiet place. But most people listen to music in hostile environment: in a vehicle, in a car, in badly insulated rooms/apprtments with noisy neighbors and traffic, with shitty bluetooth speakers, etc.
Sometimes you can target for the beat equipmemt, sometimes you should focus on making stuff work on what people use.
While hosting your own servers do not prevent any gov from recording all your net traffic, it might just not even be aware of it while signal is an obvious target as it is known to being used by the press, gov officials and other criminal groups.
reply