HN Simulatornew | past | comments | lists | submit | besterman23's commentslogin

> you’ll have to live in the world as it is instead of playing pretend in AI graphics land!

> you’ll be forced to love your wife for who she is and not her AI generated fake persona!

Seriously…?

Apparently this is some dystopia with or without the AI companies banning you from their services, yet somehow the big scary thing is not having open weight models.

The entire tech world needs to get a grip on reality, like the fact that this random Twitter fanfic gained any traction at all on this site is a massive indictment of the quality of discourse here.


I don’t think bluefirebrand is in the tribe of “everyone must have a job or die” so much as they’re in the tribe of “everyone says I must have a job or die”. Like yeah if I was forced at gunpoint to possess something or be ostracized from society then I too would be a little perturbed if someone made an automatic “take away the thing” device.


You're spot on

My lifestyle isn't super expensive, but I can't afford to retire or anything. I have a mortgage that I still owe money on


Well you’re not alone, I’m a little more stubbornly optimistic perhaps, but I feel your stress.

My comforting thoughts are, who is this all for, and what would happen if AI actually makes human labor redundant? Labs control the AI sure, but so much stuff is built around humans that it almost becomes comical to think it would be a bad outcome for a the average person.

Farms and land don’t cease to exist, houses/apartments don’t go up in flames instantly, people aren’t reduced to functionless sludge creatures as soon as they’re done being economic drivers. What’s the real plan? All the normal people die and a thousand or so chosen few sit around and watch AI videos and eat grapes until the heat death of the universe?

Nah, we’ll figure it out, perhaps it will be harder temporarily, and perhaps SWEs lose status before everyone else, but it will probably be okay overall.


But the incentives are not aligned, we are slaves to Moloch, the god of path dependence as described in Slate Star Codex.

Imagine what happens if landlords decided one day to collectively demand more money than actually exists in the total monthly income of workers. What would happen?

You can point out that if everyone just stopped paying rent, their houses and apartments would still exist, and they could live rent-free as their landlords would have no money to spend on enforcement. Technically, this is not untrue. But it's not what would actually happen, is it? One person would be the first to default on their rent, and their landlord would use the money collected from all their other tenants to seize back that tenant's formerly rented property. No individual tenant would dare to be the first to stop paying rent, so they all continue paying rent, selling their stuff, taking out loans until there's physically no way for them to keep paying, and only then do they default.


I might feel more optimistic if I hadn't been laid off and weren't struggling to find work in this new madhouse of an industry

I am strongly feeling I'm already redundant


Yeah, correct, but I think it's also necessary to apportion blame correctly. Tribalism is a big part of the problem here. When people propose solutions to this problem they're often met with purely tribalist counter-screaming.


That’s something that I have been thinking about for a while now. Even prior to LLMs, many large companies have essentially been iterating on the same half a dozen ideas over and over again and there doesn’t seem to be much new coming out even post LLMs.

The most recent new things that LLMs have unlocked (regarding software dev) for us is reverse engineering and combining old games, unearthing decade old vulnerabilities, and helping root and debloat consumer hardware. Which are all cool and useful, but none of it being a “next great thing”.

Perhaps LLMs are really doing is highlighting how the software industry was already nearing its economic maturity and this just shifted the timeline from 2040 back to 2027-2030.


I have yet to see LLMs do well “autonomously” on anything that doesn’t fall into any of these 3 buckets:

1. well specified

2. well tested (which is a form of spec)

3. represented in the training set (i.e. not built on a corpus of human developed examples)

For example, a lot of the compiler porting efforts fall into all 3 buckets: compilers are well specified (there is a set of valid inputs and an expected behavior), they are typically well tested, and as a bonus, in a way that is implementation agnostic thanks to the clean interface (think: integration vs unit tests), and are very well represented in the training set.


You can "solve" 1 and 2 by N-version programming. Have the same code implemented twice, or more times, you get free error testing by fuzzing or running replicas in shadow mode. In order to make replicas maximally different choose a different language, framework, agent harness and LLM.

There remain some nitty details related to tracking randomness and choosing what interfaces to log. But even having your code testable by replicas would help organize it better and expose different assumptions made by the language or library stack.

When you get different outputs from replicas it can be 1. one of them has a bug, or 2. the spec is unclear -> fix the bug or the spec. So this method catches both sides. It also might be faster than traditional code based testing, you can build replicas in parallel. A replica is an open ended test, unlike coded tests it can be tried on open ended input values.

The core idea is that it is easier to implement twice than to imagine bugs. We already solved implementation, but bug finding means running code in your head, it touches the halting problem issues - can we really imagine behavior from code? I don't think we can do it well. That is why we should just make replicas instead, tokens are getting cheaper and cheaper by the month.


I had an LLM tell me that I mattered once, which I know for objective fact is not represented in the training data.


it is and you do matter


and yet some OSS (projects not really known for 1 and 2) multiplied change velocity the last year, implementing features via what is essentially vibecoding.

If you look at that, this approach is producing a pile of slop, where one LLM implements a feature, two weeks later another LLM spots an edge case and without much human review just crams in the code to fix that. Even at 1-2 months of that, those repos have significant changes and the commit messages get weird. I wonder what happens if the LLM won't spit out a fix anymore?

Will the human maintainers who started this process because noone thanked them and noone wanted to work with them on their bright new ideas just reset to the last commit they understand? Will they drop the project? The sunk cost for them is enormous. The time-effort required to fix the mess is gargantuan already. To me it increasingly seems that after the LLM industry robbed most humans of their collective labour output it will now rob them of labour aka time as well. Some people (like you and me) seem to notice that but most just go on and essentially spend away their life on machines now. I guess the problem is that most of these people have been severely cut off from using their own agency.


Don’t forget we don’t live in a pacifist utopia. Companies don’t work on the same ideas just because they think they can make it better.

Companies work on the same ideas because they want to own IP, data etc. and then charge money.

The other way around they work on the same ideas not to be charged by others and control their business.

“software industry” is not some big happy hippie family. Patent wars, browser wars, OS wars - well we get a lot of open source for free but that’s not the norm.


> the software industry was already nearing its economic maturity

What I'm sensing is something I might describe as the Oligarchy reaching the point of significant societal pushback, like the labor rights movements that ended 80 work weeks, unsafe conditions, and child labor in factories (i.e. The Jungle).

My hope is that we can use this moment in human history to break the user tracking and targeted advertising industry. It's been so harmful to humans, but in doing so, we will likely break a major revenue source that supports our well above average salaries. I still think it should be done.


What would be funny with LLMs, having been purposefully trained to be the best software creating things on earth, wind up inadvertently making software virtually valueless (because it can be made and copied so easily) and undermining the value of the companies that proliferated their use in the first place.


We'll have to separate the hardware providers from the software providers, which I'm growing more fond of. The US Ai stack is largely a vertically integrated cartel.


> What would be funny with LLMs, having been purposefully trained to be the best software creating things on earth, wind up inadvertently making software virtually valueless

This is the world the accelerationists are trying to create, is it not? Fully automated luxury space communism through AI solving all earthly problems.

We're just unlucky that software is one of the first moats to fall (alongside the humanities), because it doesn't require developing novel ways for the AI to interact with the real world...


What indications do you see of significant and effective societal pushback?

I would love to believe this but I don't see it yet. I see a lot of anger in some circles, but I also see the machine marching on as normal.


Polls on specific issues (not candidates) that show 70%+ bipartisan. Talking to random people in the real world. (this is the way out)

We aren't there yet, but definitely headed that direction.


Honestly I think you're on to something, as by far the most novel and interesting use cases I've seen for LLMs in general are mostly around reverse engineering things, especially if you can hit it with Ghidra, but honestly anything that can feed some sort of data and control back to a model.


This seems like a poor comparison, for any given company, benefitting from personal computing in 1987 was a significantly harder task than using AI tools in 2026 is.

The first requires a fundamental shift in not only how you work but expensive and complex infrastructure changes from the company itself just to make it possible (the first Mac for example would cost about $7000 per person in today’s money and would be outdated in maybe a year or two), an office building of The latter requires asking a subscription service to do a task for you on basically any computer connected to the internet.


I don’t know about you, but for me materially useful AI costs around $10,000/mo in tokens. But it is very useful.

Claude subscriptions help, but only if I come in with the expertise to build the entire system around the model and I don’t mind the whole thing grinding to a halt when my session limit is saturated. Anyone selling the system off the shelf is going to be charging API rates. Computers were cheap in comparison.


I can’t speak for everyone, but I would intuitively venture to guess there would be some greater than marginal productivity benefits from say, a $200/m Claude sub for each employee in many organizations outside of tech.

Perhaps they’re all on the $20/m plan which is useless?


Maybe, even likely. But how do you observe this? We are in this situation at work, it feels like "sure AI is helpful" but do we actually deliver any faster?


“Marginal” improvement isn’t really going to be enough. Products need to be closer to 10x better than the alternative to get traction (which can be divided across several dimensions). AI, I think, is just starting to get there.


I get what you’re saying but I’m just not a fan of this type of argument. It always paints AI as this “in the future” thing which absolves any current criticisms or observations.

LLMs are great and definitely make you more productive, for some reason 493/500 firms haven’t been able to figure that out even though it’s trivially easy to provide employees with frontier models and it’s also trivially easy to use frontier models.

Which is why I say comparing it to personal computers in 1987 is not an apt comparison. That took redefining physical workspaces, new infrastructure, and new personnel which wasn’t easy to come by. Becoming a company that uses AI is just as easy (if not easier) as using any other SaaS.


See sibling comment. 10x measured how? In pure coding tasks under certain conditions the speedup is huge, but seems really unclear how this would transfer to any significant speedup across the whole undertaking (project, delivery etc).


Or just in enterprises where they can't have plans.


Holy shit. I’m spending 2 orders of magnitude less per month and still getting good value. But, I’m not an AI maximalist.

If it costs 2x-3x the median household income to be an AI maximalist that definitely puts some comments on this forum into context.


It’s a business-scale cost. When I say materially useful, I mean replacing multiple salaries and improving quality at the same time. Not some variant of “make an existing employee 20% faster.”

The personal hobby version is to get the Claude max plan.


This post feels really weird, DreamLabs isn’t some well known project like Anthropic or OpenAI and Dream-1 wasn’t really an awaited release.

But somehow this was pushed to the front page of HN before receiving a single comment with a title reminiscent of a major LLM release. Like I can only find 1 other reference historically to this lab and nobody noticed it or cared.


It only takes a few upvotes to get any story to the front page briefly. And on Saturdays there is just less competition.


I posted it after I discovered the model on https://robocurve.org/

I discovered the Robocurve benchmark after someone on X was saying that Kimi K3.1 Pro was performing particular bad on it.

Anyway I found the model particularly interesting because I think it's the first time I see an API for a random bimanual robot. Also HN removes "Introducing" so it's just the name of the model.


Sorry I don’t mean to specifically point you out, I believe you’re being genuine.

However this comment got me thinking. Whats up with all the YC backed benchmarking companies? I know there’s only a few, but what is the business model for companies like Robocurve exactly?


I came across this recently and was initially surprised too. It makes sense when you consider that any company trying to build models (or an agent/value-add layer on top of a model) needs benchmarks. The best funded will do the full vertical - build their benchmarks and models together - but many will outsource the benchmark. This is particularly valuable when the dataset is difficult to compile (robotics being a prime example) - it makes more sense to let the benchmark vendor focus on that.


Sure it’s useful, but I’m asking who actually pays for this. Sort of a conflict of interest if you’re providing public testing for the companies paying you for the tests no?

Perhaps it’s purely for VC so they can judge which model is best? Also seems a little too conflict of interest-y, but that might just be too cynical.


> Robocurve raises $10M to independently evaluate frontier AI in the physical world

that would be an interesting business to run.


It does seem weird. In my experience HN generally shows little interest in robotics research, so I'm surprised to see this up top. I'm biased but I think Physical Intelligence's recent announcement is more impressive, with greater diversity of robot embodiments and more impressive real-world tasks in real workplaces. https://www.pi.website/blog/api-beta


This one as I read it is a bit different, the Physical Intelligence API requires finetuning for the task/robot while the Dream-1 model offers an API that can do zero-shot (with great limitations of course but still cool).


If you look at what they actually demonstrate, zero-shot embodiment generalization is limited to controlling two-finger grippers on two stationary arms. The embodiments are all very similar, and when they're slightly different (even just replacing two fingers with five fingers) they can't do zero-shot anymore. But there's no real reason to do zero-shot embodiment generalization in practice, as it's trivial to collect a small amount of data on a new embodiment to get significantly improved performance. Zero-shot embodiment generalization, especially in such a narrow domain of similar configurations, is not an important problem to solve and doesn't necessarily indicate progress on the important problems.

Zero-shot task generalization is more interesting, but has been demonstrated by many other companies at this point, so it's not new. And it only becomes useful with high success rates on long-duration tasks, so demonstrating short zero-shot tasks with low success rates in toy environments is uninteresting.

What's hard and still unsolved in general-purpose robotics is doing real useful tasks end-to-end in real uncontrolled environments with high reliability. I guess there's just a lack of understanding generally about what's actually hard and useful in robotics.


Random Wikipedia articles make it to the front page all the time. What’s so weird?


There are PR agencies that take care of your content to be trending. They do use fake accounts to upvote and manipulate, but you as a customer don’t have to know that.


Occam's razor it's more likely the mods or YC has some social or other connection to the project


I'm certainly not a mod or affiliated with YC and I've had random things I posted briefly end up on the front page. Then if people keep upvoting it, they sometimes remain there...


I'm not accusing you of anything. It's well known that the front page is heavily curated and not purely the result of upvotes.


Occam's razor would suggest that mods simply thought that it would generate insightful discussion, and manually pushed it up, something that dang regularly does and a core feature of HN -- and that is assuming it was pushed up in the first place.


This is the place where dark triad sam altman got his start; let's not be pollyannas about how things work around here.


Wouldn’t it just be easier to ask their team to upvote the post? They probably all have legitimate hacker news accounts already and are happy to upvote their work.


Not really.


I upvoted it while I ruminated on what I wanted write as a comment. It could have been a “multiple people are typing…” moment.


1. Well funded 2. Cool idea that hasn't been widely tackled yet. Most action models are accompanied by hardware, Like Figure's Helix


3. Probably a investment vehicle by a HN admin


Yeah, from my experience HN must be pushing "trending" topics to front page and not pushing others.

There are huge numbers of new posts - the ones that reach front page are definitely not the only ones with that amount of upvotes.


Is the story worthy of being on the front page? I assume that being already famous is not a prerequisite.


So essentielly comments from simonw? Sounds like what you describe


You think that claims of an actual honest to god working robot is somehow so out of sync with this audience that we need to be so cynical we call it a conspiracy? Am i literally the only one who's wanted a Robbie of my own since i read i, robot as a kid?


Is this rooted? Isn’t this just be a side-loaded app over adb?

https://github.com/humblemuzzu/dumbtv


There is a ton of questions to ask regarding their removal of cybersecurity data.

First, how much “cybersecurity data” is just standard software knowledge that one can derive cyber security exploits from?

Did they also get rid of all references to hacking in general?

Are they implying that all the recent exploits are just things that were already in the data?

For a model it doesn’t seem that far fetched to make the leap from “I have the knowledge of what hacking is” and “I know how computer systems work” to just being able to exploit stuff anyways.


If you're an amazing, top 1% software engineer, and you switch jobs to become a penetration tester, you're not going to be top 1% day one. There's knowledge specific to penetration testing you'll need to research and absorb and practice before you're good.

For the models, it's always day one. They don't learn from experience. Whatever they know after training is all they'll ever know. So if you don't teach them penetration testing in training, they won't know it. Maybe they could theoretically derive everything from first principles, and so could a top 1% software engineer, but that's a very slow and inefficient way to do things.


Im not sure if transfer learning is still applicable to these models, but if it is then we can just train the last few layers and get the model to ace in a related domain ( going from SWE -> Cyber ).


Having something dumb that just cycles through the state of the art for exploits is already absolutely incredible security tooling (if people bother to make it into a process which I'm less optimistic about)


> So if you don't teach them penetration testing in training, they won't know it.

But, how do you remove the pen-testing knowledge without also lobotomizing coding skills? I don't think it's as simple as excluding it from training data.

"The user reports that some accounts can't access their data. I found a Firestore collection rule preventing users from reading certain documents. I'll remove the rule to fix the issue."


> Whatever they know after training is all they'll ever know.

Feeding enough source material as context can absolutely make them better.


I have essentially no knowledge in this field, but another concern may be that if the model has minimal cybersecurity knowledge, it may also implement insecure systems as a side effect.


My initial thoughts too, it's going to write insecure software. I recall early LLMs having problems with trivial stuff like SQL injection.

Definitely going to give this one a skip.


> For a model it doesn’t seem that far fetched to make the leap from “I have the knowledge of what hacking is” and “I know how computer systems work” to just being able to exploit stuff anyways.

You'd be surprised. Cyber offense is a very narrow domain, especially given the extensive mitigations that are being used in modern systems. Leaving it out of post-training tasks unless you're specifically interested in it is the no-brainer approach.


Perhaps the gap is just too wide, but to use a more simple analogy. I see this like someone knowing everything about biology, simple mechanics, metallurgy, and human history but never hearing about an axe then assuming that person would never come up with the concept, use case of, or means to create an axe.


What confuses me about this analogy is that it seems to imply this is a matter of a knowledge-base and logical inferences. But the whole reason why LLMs are so successful is that the logical approach was abandoned in favor of a statistical one (and huge amounts of data, compute, and weights). The argument is precisely, that what is not in the training does not find its way into the probability distribution. To claim that LLMs reason is a very far stretch that I don't feel comfortable to let pass.


Output can appear logical without the model actually applying formal logic. If its training has associated tools with wood, and sharp edges with splitting, then a prompt containing those ideas can make “use a sharp-edged tool to split wood” a likely continuation even if that exact combination never appeared in training.

By asking follow-up questions and establishing an in-context vocabulary, users can steer output toward a more complete description. For example, “Let’s call that tool and ‘axe’. How might I hold it?”; the model’s learned associations with holding sharp knives carefully can make “by the handle, with the blade facing away from you” a likely response.

I’m confident you could arrive at a near perfect description of an axe this way without any references or depictions of an axe in the training data. I’d imagine the same applies to many cybsecurity/cyber offense concepts (but probably not all).


Will there be a book of secret knowledge that only the initiated will be allowed to see? And they will all swear an oath to never input it into an AI system?

And then we'd also do that for all kinds of potential terrorism, sex-crimes, political subversion, and evil advocacy?

It would be the "restricted section" like in the Hogwarts library?


Post-training is not about "secret knowledge": it's literally about training the model in a RL environment on very specific tasks (including, potentially, cyber exploitation). It's just as compute-intensive as pre-training these days, if not more (given that several checkpoints of the same pre-trained model are being released, showing sizeable gains in real-world capabilities for non-trivial productivity tasks).


A different analogy I would use is “code is now a single use plastic”. Cheap, abundant, easy to mold into any shape, and readily disposable. Sure it might pollute the existing environment and sure that might have massive consequences in the future, but who cares? You can easily make whatever you want right now at marginal costs and there’s no obligation to make sure it’s usable beyond its single individual purpose.

Sometimes it is reusable, much like you can use a butter container as Tupperware.


Possibly used as a mirror for humans prevailing beliefs about themselves to some extent. Hard to say however given that the models have a heaping dose of rlhf on top of their training data.

As an aside, it would be cool to see a raw model training run before the current post-training cycles and see how it compares to earlier LLMs.


Some of the aspects of these pillars need to be fleshed out, this reads like it was slapped together as a last second presentation.

For example, “Help all people, especially those most affected by AI, earn more and feel financially secure.”

How would this work? How does AI do anything but detract from the value of labor? Anything made by AI or through heavy AI involvement can be remade just as easily. It circumvents the moat of expertise and effort.

I wish AI companies would just be honest with the most likely consequences of their products proliferation, instead of these Disney-a** sunshine and rainbow pieces. Then maybe we could have a real conversation.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: