What does any of that mean in practice? it's just rambling about abstract concepts that seem to be designed to hint at a bigger picture, when it's just getting AI to write code for you.
Is this where it's going? Having to mystify our roles so it seems like we're still the thought leaders when actually we're just becoming pseudo-teachers that try and herd our group of AI idiots to the right conclusion for us so we don't have to, without ever giving away that it's just all techno-babble?
This is not rambling and it is not abstract. The content here is about the second-order effects of "getting AI to write code for you" in less supervised ways. I'll cede the author could have polished this to more concise effect, but at the level it's at, a reader's failure to understand the substance doesn't imply there's mystification going on.
I agree. I think the comparison made elsewhere in the thread to Yegge's recent AI delusions/projects is especially unfair.
Armin is actively grappling with the implications of AI-produced code on the software engineering craft, and trying to reflect on how to responsibly adopt the best parts of the new world. He's recognizing that the AI skeptical/cautious are still massively impacted by everyone else using tools that run or are built this way.
I really appreciated the piece, and I'm glad we can still publish work in progress thoughts that don't have a clear thesis and call to action.
I'm convinced the field of software engineering is being split in two. These are real concerns and coherent arguments, that make sense for developers who have been using "agentic loops" and heavily AI-assisted workflows.
It scares me that someone can see this as "rambling about abstract concepts" while I see exactly what the author is talking about, both at work and in personal projects; the thought that the majority of people have absolutely no idea what's unfolding.
tech blogs used to read like actionable readme guides. I couldn't finish it without thinking: what am I supposed to do with this information? The shelf life of the latest and greatest is about 2 weeks in the AI space. I never caught up to the ralph wiggum loop and now I'm glad I never tried.
Practically, what this boils down to is having clear success criteria.
The harness (Claude Codex, Codex, Pi etc) keeps throwing things into the context and executing tools (as directed by the model) until the success criteria is satisfied.
The "rules" of using AI successfully are basically just the rules of any successful development team. Break things down into clearly defined chunks, make the success criteria clear and provide a way to get the right feedback on how the system is running (logs, metrics, traces etc).
Spending your tokens like some sort of primitive meat bag is passe and any developer doing that will be left behind by the true AI natives.
What you need to do this month is set up a central planning agent that creates a 5 year plan and then orchestrates teams of subagents to each direct subteams of subsubagents to fulfill your goals.
Have each team at every level inspect each other's work and authorize each agent up and down the chain to spend tokens on your behalf for the greater good.
With any luck, the US Administration will see the light and allow AI agents to open up their own credit cards on your behalf to eliminate unnecessary bottlenecks. Agentic Fintech is the future.
We need to move beyond this early stage where you give any thought to spending those tokens yourself!
From my past experience of religion at various levels I am very often reminded of borderline-cult religious meetings, and the zeal of converts repeating gnomic oversimplifications, and of how exhausting it was to try to engage with them on any topic of substance.
My own feeling is that it is totally OK to simply route around these people.
It's fascinating how many of the "keep your identity small" folks in the YC/HN sphere have lost any sense of perspective at the first sign of a technology that wanders into the philosophical realm. AI-oriented identities are everywhere.
Why is HN interested in management and team process discussion but allergic to similar topics on how to manage agents?
It’s like saying why discuss these team workflows when it’s just devs writing code. Or why use any jargon to describe workflows when it's just devs writing code.
Because like every other large social group it’s actually a collection of dozens of subgroups that have an overlapping interest
You’ll see those subgroups come out in threads like this and you’ll see other subgroups come out in different threads there is no singular version of hacker news that is always existing it is a collection of sentiments that from time to time align interests around specific topics
I’m sorry but this response is just absolutely ridiculous and is not giving anything near the respect to the author that you should have.
You’re just rambling and ranting about philosophical things and have basically nothing to say about the technical or engineering points that the author wrote.
This is a entirely emotional appeal and doesn’t actually engage the author where the author is engaging in the audience.
If you look further down thread there’s dozens of comments that are engaging with the content and not being hyperbolic about all this cyber shamanism or whatever you wanna call it
When someone is expected to be wizened and does not have the knowledge to keep up with the needs of those around them, they in turn become Shamanistic in their practice.
The speed of improvement on these models has been incredible and has outpaced the learning speed of humans and put many experts into these Shamanistic roles.
I think the operative means of addressing this is to recognize that we can only learn so quickly, but we are still called to improve our knowledge and understanding to a higher level.
Since the improvement of these models is neither logorithmic, nor exponential, we currently occupy a space in time in which the models are currently smarter on average than we are as a collective whole.
But it demonstrates that LLMs struggle with basic reasoning.
A criticism of LLMs is that they're imitating without a understanding of what they're doing and without a clear plan, so this inability to solve a simple logic puzzle is very relevant.
If LLMs didn't struggle with reasoning problems then something like ARC-AGI wouldn't exist.
You can fool an AI as evidenced by them being fooled. They demonstrably appear to be working through a problem and get fooled by the wording of the problem. If you think differently, merely asserting it is not the way to convince people that what they see is wrong.
An LLM is (effectively) just a really, really elaborate "choose your own adventure" book.
It's not "working through" problems, it's just tracing a route through an pre-defined information space. It's not actually thinking, it just does a good impression of it.
You can't constrain foolability to just the animate objects.
Foolability is based on intelligent processing.
To what degree such an intelligence needs to be developed is not defined at all (we can, for example, trivially fool a doog using one of many tricks, and a dog is hardly general intelligence).
Thus far, the inteligent processing we knew concerned animate objects. But we now have developed software/hw combos that exhibit intelligent processing. Is it enough to actually be intelligent or to be fooled?
We don't know. We do know it's enough to appear intelligence, convince people that it is intelligent, and to appear to be fooled.
But even if we think of LLM AI as mere mechanistic process with no emergent intelligence, who said one can not fool a mechanistic process? We can fool even simple pre-LLM gaming AI systems (based on simplistic heuristics) just fine.
They simply appear to "exhibit intelligent processing". The intelligence is in what created the data. It's a surface that's being traversed in a complex way. The LLM doesn't 'understand' that surface in any way, it just traces a path on it and reguritates apparent understanding. I'll grant it's eerily convincing.
'Fooling' something, essentially to deceive or trick, is defined as causing someone to believe an untruth. LLMs don't hold beliefs (neither do mechanistic processes), and they aren't a someone.
You can widen out the definition of the word but that generally makes language weaker - interestingly, semantic drift is a big issue for LLM's.
We have different opinions, and that's fine. Have a good day.
Please explain why the mechanism of the LLM generating output precludes it from being able to be fooled without using use tautologies or reducing to substrate for explanation.
If you're working for yourself paying monthly is exactly the same as amortising an asset. Personally I'd rather my business just pay $100 a month than have to deal with additional hardware and software maintenance while using a depreciating asset that is break-even after 3-5 years depending on the spec.
There's an argument to be made for being able to turn on a feature for a certain segment
Not just an argument, it's the entire point of feature flags for ui experiments which is an essential practice. Dynamic adjustment of the cohorts (or even just an immediate kill switch if it's a disaster) is required.
Adding another desk isn't "rethinking the desk". It's adding another desk with a slightly different purpose to the first desk. It's maximalism under the guise of insight.
He has an old picture of the desk against the wall. The new setup is the cluttered desk in center of the room. "Two-part desk" actually means a one-part desk, the source of your confusion probably.
> Alcohol-related harm is estimated to cost the NHS in England £3.5 billion every year.
If we look exclusively at numbers, prohibition would save money. If that's all we care about, try that out - oh, the Americans did, and it wrecked their country and filled it with gangsters, because no amount of trying to stop people drinking actually stopped people drinking, and normal people having to pretend they weren't going to drink, but secretly really really needing it and finding criminals to supply them with drink built out an entire parallel black economy and gave gangsters huge amounts of money and power.
If we're looking at saving money, maybe just kill the long-term disabled and elderly? Easy win for saving money! That's all that matters, after all.
Sure, except sugar in itself isn’t bad. It’s products with excessive quantities of sugar. Various laws restrict those, including the promotion to children:
You assume that there is a balance point. There is an unlimited demand for healthcare. Additionally the more money you give to a failing system, the worse it gets. It’s a positive feedback loop.
Nonsense. Did people already forget the prohibitionism? Did people already forget the war on drugs? I remember liberals were talking about drug decriminalisation 10 years ago, has everybody turned into a puritan nowadays?
Also, very hypocritical argument when alcohol (and gambling) are very accepted in British culture. I'd like to see the numbers showing that the few people that still roll their own cigs at 15 pounds a pouch cost more to the NHS than all the alcoholics in Britain.
Smoking ban is, as usual, Labour going for the low-hanging fruits to scrape the votes of the elderly that are likely to be swayed by these empty arguments, just like the Online Safety Act. One thing's for sure: Barry, 63, would not like if alcohol and gambling were regulated in any way.
I'm not a smoker any more, hate the things and can't stand the smoke, but I sure am glad to have left that island of short-sighted yet heavy-handed politics.
Let's not forget this is a policy that Barry, 63, wouldn't be affected by - only young people (let's say it's Nicolas, 30 ans). Barry, 63 loves voting for parties that fuck other people and make their lives miserable, but not him.
It's not journalism though is it, it's just someone's blog where they can tell any story they want, as has been the entire history of story telling. With that out of the way the rest of your post is just flanneling.
If the title was “I went to a single party while visiting SF and say some weird things” I might agree, but the article from beginning to end is written as if the party was a lens into society as a whole and indicative of larger trends.
There’s a motte and bailey thing going on with this type of rationalist writing where someone writes authoritatively on broad subjects and then when anyone starts responding to it they immediately repeats to “it’s just a blog” to forgive all of the problems with it.
this is likely to be the most interesting argument I will read today: is substack legit editorial journalism?
certainly there is no organized journalistic outfit behind it, but also, a lot of legit journalists want their substacks to be taken as facts of record.
You're learning at your standard rate of learning, you're just feeding yourself over-confidence on how much you're absorbing vs what the LLM is facilitating you rolling out.
The latent assumption here is that learning is zero sum.
That you can take a 30 year old from 1856 bring them into present day and they will learn whatever subject as fast as a present day 20 year old.
That teachers doesn't matter.
That engagement doesn't matter.
Learning is not zero sum. Some cultural background makes learning easier, some mentoring makes is easier, and some techniques increases engagement in ways that increase learning speed.
Is this where it's going? Having to mystify our roles so it seems like we're still the thought leaders when actually we're just becoming pseudo-teachers that try and herd our group of AI idiots to the right conclusion for us so we don't have to, without ever giving away that it's just all techno-babble?
reply