bme's recent activity
-
Comment on The machines are fine. I'm worried about us. in ~science
-
Comment on The machines are fine. I'm worried about us. in ~science
bme Link ParentThe reason I am phrasing it this way is because no matter what educators produce, the commercial/freemium models and harnesses are not going away. This is like sending alcoholics to hangout at the...The reason I am phrasing it this way is because no matter what educators produce, the commercial/freemium models and harnesses are not going away. This is like sending alcoholics to hangout at the bar. It will take tremendous force of will to restrict oneself to making life "harder" for themselves now for a payoff later. I cite the entirety of the brain-rotting internet as my evidence.
I think it's great though that schools are looking at building better tools though!
-
Comment on The machines are fine. I'm worried about us. in ~science
bme (edited )Link ParentRight: and this is exactly the genre of argument that I say I don't agree with in my opening paragraph. I am saying that all of it is bad. It's also bad by degrees. It is not hard, even for...Right: and this is exactly the genre of argument that I say I don't agree with in my opening paragraph. I am saying that all of it is bad. It's also bad by degrees. It is not hard, even for someone with middling intelligence to follow that many liquids stain, and that many surfaces can be stained. This the basis of the preservation of human knowledge is learnable with much effort. It's not going to be as good as modern inks, or industrially produced paper, but it can sustain knowledge transfer. Try explaining even an 8bit full adder and the physical circuitry for that to most people, and that's not even the beginning of doing anything useful with a computer.
LLMs and other sufficiently complex systems are bad in the sense that responsibility becomes diffuse to the point that no one will take accountability. LLMs are not special in as much as they lie on a spectrum. We've all effectively been told "computer says no" for years. Before computers decision makers hid behind their lackeys to avoid accountability. LLMs are special in the sense that they currently represent for the tip the same shitty spear, and a step change in capability for continuing this trend. Where I think they differ from what we have seen before is the function that is being outsourced: thinking itself. In this regard I do think the medium is the message and the medium is having a profoundly negative effect on people.
-
Comment on The machines are fine. I'm worried about us. in ~science
bme (edited )Link ParentThe problem I see with LLMs for teaching though is I think it takes an exceptional individual to wield an LLM and use it to increase their own understanding instead of letting it carry them. Ofc...The problem I see with LLMs for teaching though is I think it takes an exceptional individual to wield an LLM and use it to increase their own understanding instead of letting it carry them. Ofc someone could design an LLM experience/ harness explicitly around the learning loop so that it would take significant energy to hold it wrong, but that's not where the money is so I find it unlikely to be as widespread on the same timeline.
I say this as selfishly I use LLMs to do my work more efficiently and pull the ladder up behind me. We aren't hiring grads. We aren't hiring apprentices, we aren't hiring juniors. I don't know who is supposed to replace me when I go. Weird to see my generation (elder millennial) gap those below them just as hard as the ones they were complaining about 10 years prior.
EDIT: on discovering the mechanisms via which LLMs work, I agree there is progress there, I'm just not that bullish on what the possibilities are, and certainly don't think that we will bind our usage of them by our understanding of the underlying mechanics.
-
Comment on The machines are fine. I'm worried about us. in ~science
bme LinkI guess the problem is that there is a very specious rebuttal that says that plato objected to writing, people flipped out about calculator usage, and computers etc and we are all fine. But we...I guess the problem is that there is a very specious rebuttal that says that plato objected to writing, people flipped out about calculator usage, and computers etc and we are all fine.
But we aren't fine. I am deeply embedded in high tech stuff and it makes me exceptionally uncomfortable how specialized our thinking has become and how disconnected we are from almost everything we use. LLMs are the next rung on the ladder of delegation, and instead of saying "it's fine, look at all the runs we've already climbed, this the natural progression of things" I feel like we should be saying "we should maybe descend a little?" It is nuts how many things your average Joe interacts with that they have zero understanding of, and it's allowing for all kinds of nasty consequences (algorithmic feeds, dynamic black box pricing with real consequences, i.e. insurance). AI labs can't really explain why models do what they do, and while they are chipping away at the problem, we are sprinting towards using them everywhere.
I feel for the young'uns. I have no idea how they will navigate this without the grace and generosity that was afforded to me when I was a teen to make mistakes without something ubiquitous and available and way smarter than me whispering answers in my ear.
-
Comment on Why contacting customer service might not be best in ~life
bme (edited )Link ParentYes to a degree. But also it's very easy to say "companies should plan for contingency" and completely ignore the economics of those contingencies. Customers are price sensitive and generally...Yes to a degree. But also it's very easy to say "companies should plan for contingency" and completely ignore the economics of those contingencies. Customers are price sensitive and generally unsophisticated and so if you go out of your way to invest in contingency planning, insurance, better quality components, extra r&d, whatever. it is very hard to convince a customer to part with their money now for a benefit later (I.e. a staffed up call centre that doesn't keep you waiting with onshore staff).
There is no free lunch, and where a real market exists (I.e. multiple participants, lack of cartel behaviour) customers preferences are towards the bottom unless regulation raises the minimum bar, which is also typically not done unless the thing is safety critical. We get what we vote for, whether it be by ballot or wallet.
EDIT: on rereading your comment I'm actually not sure I'm understanding the spirit of it. So if you feel like I am responding to something you didn't write I beg forgiveness. Leaving the post up anyway.
-
Comment on Control the ideas, not the code in ~comp
-
Comment on Control the ideas, not the code in ~comp
bme Link ParentI agree that the premise of the article was wrong. That was my conclusion (at the root of this thread). My summary of initial chain: Me: I don't agree with article, AI often generates code I don't...I agree that the premise of the article was wrong. That was my conclusion (at the root of this thread). My summary of initial chain:
Me: I don't agree with article, AI often generates code I don't like.
DrStone: I too was disappointed, until I started doing X.
Me: I don't think X helps with my disappointment.To be clear: I think X is a useful collection of techniques. If it isn't clear already, I use AI a lot, generate a lot of code with it, inspect a lot of code. I am not at the toe-dipping staging :)
-
Comment on Control the ideas, not the code in ~comp
bme Link ParentOk. So point 8, is effectively the antithesis of what the proposed article is saying right? It says don't look at the code, you say "I look at the code". Planning, adversarial review, session...Ok. So point 8, is effectively the antithesis of what the proposed article is saying right? It says don't look at the code, you say "I look at the code". Planning, adversarial review, session clearing, it's all a big dance to try to get the AI to do what you'd like, but in the end you still aren't actually delegating in the way that the article is suggesting. I'll also give a concrete example of the type of thing that I see all of the time. I wanted claude to produce a network manager dispatch script to update a particular route on an interface. This is a 5 line prompt at best and definitely not worthy of the complete process you have described. I am using opus 4.8 max effort. It produced something like the following
interface=$1 gateway=$IP4_GATEWAY if [[ -n $gateway ]]; then gateway=$(<<bunch of shell to try to calculate a gateway); fi [[ -n $gateway ]] || exit 0 ip route replace 10.0.0.0/8 via $gateway dev $interace metric 100Now the problem with this (ignore quoting, action checking, it's a snippet) is basically that on a network manager managed interface if $IP4_GATEWAY isn't populated it's because there isn't a default route yet. So trying to calculate one is pointless. When one shows up the script will be called again, and so it will be populated. This type of shit happens all of the time. AIs cannot keep enough in their brains afaict to stop just jamming in extra unnecessary backstops. They do it for scripts, they do it for functions, then never lean up or unify error handling, or trace architectural invariants. They just double up piles and piles of redundant checking, which they themselves then fail to keep in sync.
All this to say, nice process, but no amount of planning gets frontier models to produce code that I am happy with, because the gap isn't in understanding what I want in terms of functional requirements. The above script would have worked. It's just needlessly complicated.
-
Comment on Control the ideas, not the code in ~comp
bme Link ParentDo you not find that docs produced by ai tend to miss the forest from the trees over indexing on minutiae to the point of rendering them nearly useless for humans (probably very good for other...Do you not find that docs produced by ai tend to miss the forest from the trees over indexing on minutiae to the point of rendering them nearly useless for humans (probably very good for other agents)?
Also, I don't know about you, but personally I find I arrive at leaner designs by stress-testing them against existing code. Sure, if it's something that fits an obvious archetype or design pattern you can probably anticipate a fair amount.
I find that AI basically never simplifies ever. Neither in design or code and without active anti entropy you end up in a bad spot, and no amount of design gets you lower entropy.
-
Comment on Control the ideas, not the code in ~comp
bme Link ParentInteresting idea. I'll give it a whirl. I tend to just roll with auto-mode because I've found empirically (literally drag racing the same problem) that by the time my colleagues have finished...Interesting idea. I'll give it a whirl. I tend to just roll with auto-mode because I've found empirically (literally drag racing the same problem) that by the time my colleagues have finished their glorious round of dialectics with some planning mode i have iterated my way to a better design and solution just by hammering on the problem and getting deeper into the problem than can be discovered by looking at from afar.
-
Comment on Control the ideas, not the code in ~comp
bme LinkMaybe antirez is some god-tier prompter, but I find myself constantly dissatisfied by the decisions that anthropic's frontier models make. I think they are 100% great at automated reviews...Maybe antirez is some god-tier prompter, but I find myself constantly dissatisfied by the decisions that anthropic's frontier models make.
I think they are 100% great at automated reviews (relatively low cost to verify, I can overrule without negotiation, endless attention any time of the day, can be fed house review checklists). I think they are great sounding boards for design. Exploiting their encyclopedic knowledge of general techniques to direct research yields better designs faster than I could unassisted. I am just struggling with the idea that you can be so sure that your ideas are being faithfully reflected if you look never. That's not at all my experience. I am constantly finding inconsistencies and stupidity any time I take my eye off the underlying evolution of what's being spit out.
-
Comment on European Parliament achieves upgrade to air passenger rights in ~transport
bme Link ParentLet me get this straight. Ryan Air's customers who are incredibly price sensitive, will suddenly stop flying with Ryan Air if they aren't forced to use the app to receive their boarding pass,...Let me get this straight. Ryan Air's customers who are incredibly price sensitive, will suddenly stop flying with Ryan Air if they aren't forced to use the app to receive their boarding pass, which is achievable at near zero marginal cost to Ryan Air (i.e. won't affect their prices). Ryan Air isn't retaining anyone for any other reason than it will transport them to where they want to go cheaper than anywhere else.
-
Comment on European Parliament achieves upgrade to air passenger rights in ~transport
bme Link ParentBut there is marginally zero cost to this. It's software. It's the easiest thing on the planet to send a boarding pass via email in six different formats. I could shit this out in my sleep with...But there is marginally zero cost to this. It's software. It's the easiest thing on the planet to send a boarding pass via email in six different formats. I could shit this out in my sleep with one hand tied behind my back. It is significantly harder to maintain a complete application and keep in on multiple mobile app stores. It's not a cost thing. It's a control thing.
-
Comment on Zig creator weighs in on the Bun Rust rewrite in ~comp
bme Link ParentMy other major problem with it (the blog post) is that it doesn't pass the timing test. If it was moral and right to share all of this stuff now, why wasn't it moral and right when there was still...My other major problem with it (the blog post) is that it doesn't pass the timing test. If it was moral and right to share all of this stuff now, why wasn't it moral and right when there was still a salvageable relationship? I can believe for instance all the information is good and true, but if there was some imperative to share it, it shouldn't have been constrained by convenience if you want to claim some kind of principled approach to this stuff.
Shitty behaviour, even if I do in general like zig and Andrew's ideas.
-
Comment on Arch User Repository compromised, 1500+ packages affected in ~tech
bme Link ParentEveryone using even a remotely competent AUR helper sees the diff of every package automatically by default. It is reasonable, I did it for a over decade. 99% of aur diffs are just hash + version...- Exemplary
Everyone using even a remotely competent AUR helper sees the diff of every package automatically by default. It is reasonable, I did it for a over decade. 99% of aur diffs are just hash + version bumps, and any that aren't deserve scrutiny. The point of the AUR is no permission, no vetting no control. I packaged things I thought were useful and if it has been turned into a full time job by some kind of background vetting process I wouldn't have bothered.
In some respects I welcome this, not because I think the AUR should change but because it will force people to wake the fuck up and confront what they are doing. Do you want to be in charge of your computer or not? If you don't, buy a mac and let Cupertino decide (or Redmond / windows, I don't care). If you do and you want to use the AUR, read the diffs. If you don't then use something else. This isn't deflection. You can only make the experience of using a computer so safe before it becomes a phone. I don't want a phone.
Ubuntu has app armour, fedora ships with selinux ootb. Flatpak has its sandboxing, Wayland has completely locked down the compositor. Landlock, bubble wrap, namespaces, systemd service hardening, the whole trend is moving towards better security, and better security primitives.
-
Comment on rsync and outrage in ~comp
bme Link ParentSolid agree. When using a harness I absolutely watch and correct code and it streams through, change the architectural direction etc. Do I still ship bugs? Yes. Are there more bugs than there were...Solid agree. When using a harness I absolutely watch and correct code and it streams through, change the architectural direction etc. Do I still ship bugs? Yes. Are there more bugs than there were previously? By all measures I have for myself: no (net new bug fixes that can be tracked to generated code). Are there more bugs tickets overall? Yes, but because I have more time to push harder on our codebase things that would have been ignored previously are being picked up. I don't love LLMs, I think we are absolutely taking an axe to the talent pipeline that I don't think we will ever recover from and at this point I only hope that tools become good enough to cover for the expertise that we are losing at an insane rate as people age out with no replacements of equal experience.
None of that means that Claude code caused an increase of bugs per loc in rsync.
-
Comment on Slate Auto to offer modular electric vehicle pickup for $20k with EV Tax Credit in ~transport
bme Link ParentMost charging should take place at home (assuming you have a driveway / dedicated parking). Where I live that's absolutely true. No grid upgrade required, it can all happen over night at a rate...Most charging should take place at home (assuming you have a driveway / dedicated parking). Where I live that's absolutely true. No grid upgrade required, it can all happen over night at a rate that's kinder to your battery and can be coordinated with your energy provider to allow the most amount of charging to happen on the limited infra that we have, allowing for a smoother and more realistic grid upgrade cadence.
It isn't feasible to stick fast charging everywhere. It's bad for the grid, expensive, and leads to worse outcomes for the vehicle. Most people (again UK, but average commutes in the US while longer are typically not more than the range of an EV) can use an L2 charger at home just fine. I monitor a fleet of hundreds of thousands of L2 chargers, most people are charging twice a week overnight. 90% of the fleet is idle at any given time. If you are doing hundreds and hundreds of miles a day I would say an EV is currently the wrong fit, and at most a hybrid makes sense.
-
Comment on Slate Auto to offer modular electric vehicle pickup for $20k with EV Tax Credit in ~transport
bme Link ParentBut $50k is nowhere near the cost of installing a fast charger. A nice model from alpitronics is more like 4x that. Then there is the site cost, and any extras for grid interconnect upgrades. 4...But $50k is nowhere near the cost of installing a fast charger. A nice model from alpitronics is more like 4x that. Then there is the site cost, and any extras for grid interconnect upgrades. 4 stations is multiple millions.
-
Comment on Utah's shrinking lake: a scientific asset and a crisis in ~enviro
bme LinkI didn't realise ~enviro was twinned with r/atheistcirclejerk.I didn't realise ~enviro was twinned with r/atheistcirclejerk.
I spend an embarrassing amount of time talking to Claude and the whole essay screamed Claude to me. I didn't want to say anything because it felt like turning the discussion into a bun-flight about authorship was less interesting that the discussion about the underlying arguments themselves.