HN user
if you wanna be friends bumps-anima-0q@icloud.com
A big fraction of senior executives' work is vulnerable to replacement by AI.
what? how? literally what work is already not done by some subordinate that an executive does now?
if some work can be replaced, its already done by some employee. those employees are susceptible to replacement.
This is unusually reactive response to normal language.
Why do they need prior consent? What sort of rent seeking do you want?
The lesson of the last 50 years of the computer and software marketplace is that free and low-end eventually wins.
Mac has won and it is not free.
Iphone has won and it is not low end.
sanest comment here
umm do Mac vs pc and Iphone vs android
Commoditising complement
Sunk cost fallacy is not relevant to future investment.
investors should be cautious about
if they are cautious, what would make them invest in newer bigger models without the expected return? generosity?
To everyone praising Open weight models, could you answer a simple question?
If Anthropic doesn't make money because of distillation attacks, how would they convince investors to invest in them, such that it makes financial sense for Anthropic to train even bigger models?
Assuming it is preferable for everyone that we get better models in the future. Distillation attacks remove the financial incentive.
I completely agree with you except the environment part which seems forced. It is valid for people to be concerned with their jobs. But in writing this obvious thing we both agree about, you have forgotten the more interesting part: Waymo stands to reduce a significant number of deaths. What does it say about the ethics of labour unions? In fact, if you could leverage any control over all this, what would you do? Release waymo or protect people's jobs?
Who do you think came up with the term Luddites?
You are confusing a lot of things here. Like the existence of the movement and the people itself. And you are questioning the existence by using an adjacent topic of Luddites were not against technology.
Luddites did exist. And the people opposing data centers now are the same type of people.
Classic case of Motte and Bailey applied.
This is a disagreement on hiring Java specialists vs generalists and not about using Java in Google.
Its an interesting question and I'm biased here. As for the morals of both parties I'll leave you with this: Waymo is cool tech that actually works and saves lives everyday. These are the parties lobbying against Waymo
- Trial Lawyers because they gain from accidents [1]
- Labour Unions [2]
- Transport Unions [3]
Most recently Uber also opposed it but in a different format, but still the same. I wonder what your take on this is? Who is being anti-human here? Who wants to help people? And who are selfish? For me it looks like each one of these parties would rather have less safety in streets than have their jobs taken away.
[1] https://marginalrevolution.com/marginalrevolution/2026/07/tr...
[2] https://teamster.org/2026/03/illinois-teamsters-demand-lawma...
[3] https://www.twu.org/waymo-testing-in-nyc-is-a-danger-to-the-...
Interesting, do you think the current form of anti technology ideology is different from Luddism? In what way?
Do you mean to say that the current protests are actually against technology for the sake of being against technology?
Also, I don't say that they are against technology for technology's sake, but the real actual reason they are against technology is because they are fundamentally conservative. They act in their own self interest as normal people do and try to conserve their status, employment and lifestyle. Technology tends to act against it - naturally they oppose it.
start sponsoring the local high school gymnastics club
I would prefer using taxes instead of one-off theatrical gestures. And it looks like data centres already pay tax. Maybe what you want is higher taxes? Possibly a good idea if it doesn't disincentivise the construction in the first place.
It is completely organic and it is not new at all. Every step change in technology involved people like this (though this one is a bit unique in size).
This type of thinking: "slowing down", "being in harmony with nature", "be content with what you have" is the default mode of thinking for most rich/educated people.
(Note: this type of thinking comes at the upper echelons of Maslow's needs - impoverished third worlders don't have takes on it)
What I think is that most have had enough of chaotic progress and they just want some inkling of stability and some breathing space to even contemplate on the last 20 years of change. What they are instead seeing is a barrage of stuff thrown at them one after the other. I can understand them. And so should all of us.
The suggestion that this is a CCP play shows a lack of theory of mind.
I don't understand your point. How do you make your future code resistant to bugs without LLMs?
I think the mechanism is that ACM has a certain archetype of "bro" self selected into the group. They all have similar views like hating big tech, loving FOSS, pro government regulation and so on. The author probably fills a gap of a priest. This publication itself needed a priest to do the sermon and who better than the author. He touches on all the points of interest here and there and keeps everyone satisfied.
At least, this is what I have come up with because this blog is mostly incoherent blabbering.
At which point we will continue on as usual.
This part is the load bearing claim. Why would you continue on as usual? I'm using LLM's everyday on code reviews and they still catch bugs.
He really seems to think he made some deep insight on the "bubble" that he seems so confident about.
So, it is not obvious to me who will be training new iterations of these models once the current bubble explodes, in particular if the returns are diminishing the way I have experienced.
It looks like he has no clue on how market equilibriums work. He really seems to think LLM's will just like.. stop existing.
So in their world, people would suddenly realise that AI is actually not that economic and we can't have Opus 4.8 quality models just with updated knowledge cutoff perpetually. So in his future, things won't just stall, they will literally go back.
He's really putting his emotional weight on this particular kind of future.
Either that or he's making nebulous emotional claims - its his blog so he can do it.
Is there any proof that they are not good at special reasoning? Arc agi 1 and 2 are saturated.
I disagree. From what I see, company specific stackoverflow like mailing lists or slack channels are also dead.
The normal day to day devs just don’t have the need to go to stackoverflow anymore.
One would have to explain both consequences or dismiss it as coincidental.
FWIW I rarely have the need to ask questions at the programming level to anyone anymore. It’s just not the type of thing I bring up or anyone else. We now talk about architecture and company direction.
The company my friend works for has a slack channel for help with code, like an internal stackoverflow. It’s almost inactive now.
I asked to see one of the questions from 2024 - it could have been solved with one LLM search.
We have eliminated a whole genre of peer to peer communication.
nah i'm not cherry picking. its a serious artist and director betting on AI. expect more to come - just like we did with photography.
notice how you didn't give me a metric that proves efficiency/inefficiency?
This doesn't mean anything. Here are the facts
1. models themselves are getting cheaper - around 5000x in the past 1.5 years
2. ironically, if it were fully cheap, the same crowd would say that this would make AI bubble pop because where would they get moat?
3. training is also getting cheaper, the whole lifecycle to train and do inference on models from 1 year ago is on the whole 50x cheaper or so
I also think the author's point is a bit nebulous
Chatbot companies are aware that their products are inefficient. Some have found techniques for improving performance, but they have not yielded significant gains
Performance has increased. To you specifically: what metric would falsify the claim that "they have not yielded gains" and "their products are inefficient"?
The same metric should apply to
- industries like steel
- pharma
- internet/cloud computing
- automobile
- consumer electronics
All of them have had high impact. So any criticism on inefficiency should be unique to LLM's and shouldn't apply to all of them. Please answer.
The two claims are not related, my bad for making it seem so. I about that a serious person would still needs proof that AI is becoming more efficient. If you need proof, here's one:
A year ago, we verified a preview of an unreleased version of @OpenAI o3 (High) that scored 88% on ARC-AGI-1 at est. $4.5k/task
Today, we’ve verified a new GPT-5.2 Pro (X-High) SOTA score of 90.5% at $11.64/task
This represents a ~390X efficiency improvement in one year
https://x.com/arcprize/status/1999182732845547795
This is not even including newest improvements. GPT 5.6 beats GPT 5.5 by 1 OOM.