3,824 karma · joined March 10, 2017
I don't know if I am the same. But I do find it very difficult to use an internet browser with no adblocker.
Perhaps it's because I have now spent more time using browsers with blockers, than without
I thought that was the whole motivation for building HTTP/3 on QUIC (UDP)?
It's anecdata and maybe, MAYBE, a spreadsheet. Or a Google Form somewhere.
I've found that the LLM generated Go has few mistakes, and generally isn't too obscure. But the volume of code is so high, colleagues do a bad job of reviewing it.
I've seen a lot of very silly decisions made, like returning the wrong HTTP code, or miscategorizing a metric used for an SLO, that I just don't think is helped by the sheer volume of code one has to wade through.
Ironically, we are considering migrating some initiatives to Rust, exactly because experiments indicate it works well with LLM development.
Cows are the most common animal on the farm, and overall they produce the most value. But they are very replaceable. When times are bad, you slaughter them.
Cows are like your rank-and-file application developers: you scale them up and down with the times.
Goats are more niche. You don't have many of them on your farm. But they solve important problems (eating weeds), and they look after themselves, so you seldom slaughter goats.
Goats are like specialists in your company. People who know how the "real" things operate. You don't need many of them. But they are involved in enough critical things, niches that can't be scaled down, that you seldom lay them off
If you just mean a large company, I would say they do exist - though it may take some looking for them.
You can find them in companies that have to deal with "real" things (hardware, factories, production lines), or where there is an interest in taking advantage of emerging technology (advertising, e-commerce)
I would call my current project relatively systems-level too, as it's a network proxy. Not quite kernel level but definitely not trivial "if this then that" style coding.
My perspective is that application programming - CRUD, forms, IO orchestration - was always vulnerable, even before AI. Think about APIs for payments, APIs for subscriptions. E-commerce in a box type solutions.
That's why I always pushed to do more systems level work, on more exotic or weird technologies. It's not because I think I'm a better programmer, than someone slinging Spring code or React forms. But because in this industry it's better to be a goat than a cow.
It doesn't feel good, but at some point we have to ask, can we support every single case, at the expense of the whole? I'd like to be convinced that's not a necessary tradeoff
The appeal of digital systems, at a social level, is that they can transmit information accurately for almost no marginal cost. They come at the risk of fraud.
With AI the levels of risk have increased. Anyone can now be a fraudster, and it gets harder and harder to prove anything online. Either of us two, in this conversation, could be a well tuned LLM.
So society, collectively and at an individual level, is forced to reassess reliance on purely digital information. Which means that the advances in the technology, actually end up reducing the utility of digital systems more generally.
You can argue that this makes education costly and less accessible. It does. It absolutely does. There is also no alternative except to see higher education get wiped out, which would have devastating economic and social effects.
So we are going back to blue books and vivas.
The Unabomber (Theodore Kaczynski) killed 3 people and injured 23 others across 16 bombings between 1978 and 1995
It's absolutely incredible to see this talked about as a "reasonable" position on AI taking jobs.
edit: updated as it is not clear if Mosura is being serious or exaggerating for rhetoric
Secondly, we already know what happens when sectors face mass unemployment from the 1980s, when steel and coal workers were laid off en-masse: nobody came to help
Accelerationists want to see their jobs go up in smoke because they think that if enough white collar workers are impoverished, everyone who was formerly poorer than them, will chip in to rescue us
Just like they didn't in 1985...
In a context where the state already has less tax revenue due to mass unemployment! It's completely bananas thinking that only survives on internet brainrot forums
The only reason to emulate human intelligence is to capture it, exploit it, and reduce the negotiating power of those who have it naturally. Anyone who wants this is not a friend of human beings
I am not calling you out personally here, but this comment bears many hallmarks of the "accelerationist disease": thinking in abstract models rather than in concrete terms; optimising for the fangs of capital rather than human beings, imaginary infinite bets that lead participants to nonsensical conclusions
You can't capture the full range of behaviours and edge cases unless you are willing to ingest billions of events without lossiness
Metrics have limitations of cardinality. Logs need to be reduced or else you are in archive rehydration hell
There will be code paths that have never been triggered that you still need to reason about as part of analysing a system, eg in threat modelling
> I get why you think I might not have worked on large systems
I didn't, necessarily. But it looked like an excessively strong claim, combined with a challenge to the author's competence. I wanted to call that out.
> This is complete lack of experience on the authors part
Actually I'm going to turn the accusation back on yourself: if you think reasoning about a system's entire graph of behaviours from runtime traces is trivial, I don't believe you've worked on a truly complex system
The range of what a program can do is much larger than what a program does do in your two week Datadog aggregation window
- Type system is too concrete, can't express sets or unions.
- Not much support for functional programming except passing closures
- Can't express immutability. Well, there's const, but it's crippled
These things mean Go falls short of what GP wants.
But Go has very distinct and (honestly kinda weird) design goals, it isn't really supposed to be "the simple applications language". It's supposed to be "C with NewSqueak", very much a systems programming language. You see that in its aggressively concrete type system
Codeberg wants to protect their hosting from the large volume of AI projects.
What do you think the right solution is here?
> I think you're being dishonest in your framing
Asking without malice: why are you telling me this?
This is an open internet forum. You can suspect my sincerity but also, I can suspect yours.
Neither of us can ever prove we mean what we say. We just have to muddle on and ignore anything we think is trolling
Why do the very vocal "AI people" online have this thing where they are simultaneously:
- proclaiming their thing to be inevitable and massive and a bulldozer of productivity;
- responding to any negative reaction like they are almost like a persecuted group
It seems paradoxical to me. If your LLM-written project is so amazing, who cares about the critics?
Let them have their corner of the internet. There are plenty of other sites that will host your project.
Perhaps if you are a heavy codeberg user it is annoying to move onto Forgejo. I get that. But ultimately Codeberg has the right to decide these things
The issue being, gratification is rarely a good guideline. It just means collapsing the gap between doing the thing and the idea of having done the thing. But that gap is where you actually learn things
It does seem to be happening - at least in mobile app stores.
There's some recent analysis that demonstrates how, despite a huge updraft in the quantity of apps released, the aggregate count of reviews and downloads remains static.
In other words, there are now many more apps. But not many (or really any?) more users
Take a look at p40 / figure 12 of "WRITING CODE VS. SHIPPING CODE: PRODUCTIVITY EFFECTS ACROSS GENERATIONS OF AI CODING TOOLS" (https://www.nber.org/system/files/working_papers/w35275/w352...)
Their analysis is on pg42-43
> they should be clear about why this time is different to the entire history of the computer industry so far
I can't prove the pie is fixed, but nor can you prove the pie is infinite.
Maybe this comes close to sounding patronising, but I think the key thing people miss, when talking about economic growth of software is, money has to come from somewhere. Someone has to give it to you. So it you want to keep growing, you need someone who isn't paying for software, to start. Who are these people, how much money do they have, and what other costs are you competing against?
When clients come to you, is that because they always intended to get help taking their system to production? Or is it a last resort after all the AI approaches fail?
Is there a general way that these projects break down? Or are the failures usually subtle?
His work is very flimsy, and I have been a hater for close to 10 years [1].
I think Bregman skirts close to the "Effective Altruism" movement and his work has similar problems of choosing flashy, exciting, elitist projects over boring, uncomfortable, policy changes.
His enamourment with "AI!!!" (exclamations mandatory) is par for the course. Basically a fantasy that if AI leads to enough layoffs, the rest of society will accept a transition to UBI (against their own interests)
Bregman has been going on about UBI for decades and I've never seen him do the actual maths. In Utopia for Realists he argued the budget deficit can be completely made up by the cost savings of having fewer benefit systems. It's fantasy
[1] https://www.breck-mckye.com/review-utopia-for-realists-rutge...