dcow

If you're not sure why I brought up the hypothetical situation and you don't want to engage with it then... don't.

I'm curious, which is why I brought it up. In my worldview humans should have a right not to be advertised to and all advertising should require consent. Even today, humans can ignore and block ads without committing ad fraud. So I don't see a difference between humans and bots consuming ad-laced content. --because no guarantees can be made as to wether an ad will be "consumed" or not, only show

Since I don't distinguish a difference, I wanted to discuss whether the case at any level hinged on the bot factor. That's all.

What I do value is the ability for institutions to control who uses their space/services. So I support regulating and punishing inappropriate use/access. In this case that would be bot usage of Spotify in ways that breaks the contract you agreed upon when you signed up to use their service.

But there's a difference between "this person committed fraud and it's criminally bad because something was fraudulently obtained from advertisers", and "this person breached their contract with spotify and possibly committed criminal computer abuse etc". So it's semantics around who is being defrauded.

No that’s not my argument actually. I don’t think theft here is justified just because it’s a streaming service or big ad or it’s just a click. There’s a real crime here.

I am uncomfortable with the idea that something that is allowed if you pay humans is also illegal if you use bots. That feels like it deserves more discussion. And it feels like while there is a clear crime here, it’s also a problem that the ad and streaming industry have a model that reduces one stream to one human. I don’t want to worry about jail time for automating tasks with bots and accidentally falling on the wrong side of some industry’s really bad/dated monetization model and having it look like fraud. That isn’t the case here, but it encroaches.

It’s not about redistribution for me. I’m not of the mind “fuck big advertising and streaming any theft from them is justified”.

It was more about making sure we don’t punish people for something that would otherwise be legal if they paid humans to do it but instead used bots. And it doesn’t seem like that was the case here after more discussion panned out.

And also that we don’t ignore the minor but still real part of this that’s due to a really poor assumption that “one stream == one human”.

I wasn’t detailed enough. What I was actually curious about is whether it mattered that bots were used.

d.1) convince humans and they use their main account

d.2) convince humans and they use and alt account

I think the resolution to my question is: it’s still fraud if humans do it, like you said. And it makes me wonder if more things should be considered fraud like paying people with free products for Amazon reviews or purchasing GH stars.

I don’t think fraud requires enrichment, does it? I know we commonly associate the two but it’s somewhat of an annoyance IMO that we only care about scheming material misrepresentation when it involve money.

Edit: interesting it does look like there must be an object of the fraud, likely to prevent the law from being used broadly to punish lying.

It’s rhetorical. Setting aside the obvious credit card fraud because I didn’t know that was a factor when initially asking the question, no it’s not clear to me that using bot accounts to consume a digital streaming service is fraudulent or morally wrong the same way assuming a fake name to use someone else’s credit card is.

I don’t care if a service seeks civil remedy for any damages due to a breach of contract.

I also don’t fancy a future where people have to be worried about criminal penalties for spinning up a bunch of agents to automate something that enriches them.

Is buying likes and gh stars fraud? Is it okay if humans click the star button? Is it okay if a large distributed workload does it?

Anyway my actual stance is that it looks like in this case it’s not the bots that are the key focus of this matter so my earlier comment is mostly just a digression. And that I’d hope this case would be ruled similarly if for example human users were paid to stream his songs, however that would be adjudicated.

Yes the list is rhetorical. Yes I am aware this isn’t securities fraud.

Personally, I am not sure I consider pointing a bot at a service to use it, even if it enriches you, generally to be evil or fraud. I think now more than ever the concept of a bot is almost meaningless. So my angle with the comment is to talk about the spectrum and understand why people so clearly see (e) as criminal relative to the other situations when the harm or effective outcome is the same.

We seem generally okay with deception until it is used for personal enrichment. Why is that?

I find this fascinating, actually. On one hand we must disincentivize fraud. On the other hand enforcement of laws really should have observable utility and not feel arbitrary. I’m much more on the “fix the model” side of the spectrum on this one. That doesn’t mean I think we should allow what SMITH did.

I feel uncomfortable when we start handing out prison for things that are isomorphic in utility to things that seem clearly legal.

Which of these are fraud:

a) I block or skip ads in media I consume

b) 1000 people block or skip ads in media they consume

c) I convince 1000 humans to play my songs every day and all of them listen to the ads

d) I convince 1000 of humans to play my songs every day and all of them skip the ads

e) I program 1000 bots to play my songs every day

(b) (d) and (e) all have the same outcome for ad agencies, nobody hears the ads. If (e) is the only one that’s problematic, then we aren’t solving the problem by prosecuting (e).

Similar vein to Matt Levine’s “everything is securities fraud” takes.

Genuine question: what if my agents produce better output if I stream audio to them while they work? Is that fraud?

Came to share this setup too, though I’ve managed to keep the doc in the monitor.

Very neat and fun project though, and good to know there’s a nice solution if the need every arises.

No that’s not what’s being said. Nvidia gave him the wrong total. He paid for that total. He’s saying there’s a world in which he is owed the other 9k shares. He put some lawyers on it and it turns out it won’t be this world.

And even if you prefer to view the constitution as primarily setting up restrictions on government power, I think most of us here would agree that the at-scale data collection and surveillance our government performs under the guise of national security is an incredible power that is new and novel enough to warrant a discussion on keeping it in check.

The federal government a) saves everything, b) is constantly trying to attack encryption, c) tries to insert backdoors in hardware, and d) otherwise gives itself the means to spy on citizens and aliens alike. Even though we fight and have won, it’s been pyrrhic. I’m tired of constantly having the battle every time a new generation forgets why privacy is important.

I can’t imagine ways in which an amendment asserting every citizen’s innate ownership over their digital footprint, data sovereignty, etc. could be a misuse of the constitution. “No government shall abridge access to secure and private communication technology. No government shall collect or persist a profile of any citizen except as required to serve justice. No government can compel decryption of private data. Etc…”

The constitution applies to government, yes. But it's predicated on an enumeration of inalienable human rights that need to be recognized and defended. That's what makes our government different from other instances of government. The enumeration of rights sets the stage for what we value as a society. And privacy is sorely missing from the modern dialog.

I'm sympathetic to the idea that the right to privacy in the digital age is fundamental enough that it needs to be distinguished in our founding documents and elevated to a level on par with the freedom of expression in more than just case law. Most importantly, the government shouldn't be able to spy on citizens either. Right now it does. This has implications for encryption law too.

The point of an amendment is to establish the axioms upon which we evaluate all future laws. From there we can build a legal framework that applies similar ideas to the private sector, as and where appropriate.

"Hey look I got this brand new tool it can do everything my old tools did and more but I'm scared to use it to do more--might shoot myself in the foot."

I think most people have felt that way before. Only way forward is to practice with the new tool (=

> I think what's nice about these tools is that I can just write a document that says things like "prefer immutability" and then I neither need to micromanage nor accept code I don't like

Why are we even arguing then? You and I agree. Did I ever say "don't share a single preference with the AI"? You set the guardrails and preferences and the AI follows them. This is how it's always worked so it's reasonable for me to assume that people "fighting the AI" have already done this and are being overly pedantic about the output. Otherwise it wouldn't be eating up inordinate amounts of time...

> Have I always been the manager of a team of bash scripts? No.

No. You're not even remotely close here. Let's revisit this once you've figure out how to have a team of bash scripts implement 100k lines of code and build entire systems in 2 weeks based on high level instructions and requirements shared in context and prompts. You have responsibility at a different level and scale in an AI native workflow.

Is that my characterization? This subthread is about how it’s so time consuming fixing every little thing the AI does to be just like you would have done. My challenge to that sentiment is “give it some freedom, don’t micromanage it.” That’s all.

I understand the boundaries of ownership and responsibility. That’s why I can tell you if you are spending inordinate amounts of time correcting AI code then you’re doing something wrong. Either write the code yourself or reassess your ownership boundaries. You’re acting as a manager of a team of agents in an agentic coding paradigm. Managers don’t tell me how to write code.

If it works why fight it? Because it's not what you would have done? Implementer is getting blurry to me, at least.

Of course it's fine to have preferences. I think you missed the nuance. Bikeshedding implies the time is wasted because the topic wasn't important in the first place. The color of the bike shed, as it were, has nothing to do with the storing of the bikes.

I am equating the nerve that develops in people that get lost bikeshedding (wasting time on inconsequential parts of the problem) with fighting an llm on inconsequential implementation details.

We most certainly agree: what matters should always be the actual requirements (functional, security, performance, etc.) You can't bikeshed an important topic. Everything else is implementers decision. In my experience an experienced engineer understands the difference and trusts implementers to make the decisions that they do own.

I don't know what metric Apple implemented that has caused these engagement spam prompts (ads) to appear in their platform, but I want none of it. I've turned Siri off in the latest macOS and iOS because it's so damn slow. I'm bullish on local LLM compute but Apple it's not ready yet or not in the way you've integrated into the user interface! When I am attempting to search for an app on my system I don't want an AI response that takes 2 seconds to start. I want to see the filtered list of apps immediately. When I want to search a term I want the results immediately with any slow AI interpretation fed in asynchronous. It's painful as hell to try and use the new Siri and be resigned to wait a few seconds for any sort of answer. The worst part is you have to wait 2 seconds and then 98% of the time Siri is wrong our outdated and you have to redo the query somewhere else. When it became more convenient to jump into chatgpt or claude I pulled the plug on the new Siri.

My impression of you is not something that can be right or wrong. My challenge to you was that a userspace program can brick a machine. I made that challenge because you predicated your ad hominem attack on the idea that a userspace package install couldn't brick a system: "You calling a borked package install in userspace almost bricked is telling about your technical capability." You haven't refuted it, because you can't. And the fact that you missed rebutting my challenge (claiming you didn't see the point) and argued past me with a different point that video card drivers are not the same as the primary bios tells me enough to form an impression.

Anyway.

This thread spawned off of a discussion about how people still use Apple because though Linux may be easier than it's ever been it's still hard for anybody to run a Linux desktop system without occasional downtime and repair when something breaks. That was the point of the comment you initially responded to.

But, you missed it; you responded to:

"and almost bricked itself?? Not sure, had to fix apt"

with:

"well um axually you don't know what bricked means and that's probably a sign that you're too dumb to use linux. people who are too dumb to use linux shouldn't use it".

Even though we agree on the point, I felt like your response wasn't fair or really even apropos. So I replied. To be crystal clear:

1. they did fix their problem, but it was not a fun time for them. I can empathize 1000% with that. Sometimes I just want to do something and Linux (and everything I sign up for when deploying it) itself gets in the way. When an apt update causes a boot failure, irrespective of whether it's a failing driver or a borked firmware, it's annoying and it gets in the way. It's not an issue of technical literacy. You still have to figure out the issue and fix it which costs time and attention that is so scarce these days.

2. "bricked" has evolved colloquially to refer to a spectrum of failure modes covering everything from "the firmware is toast" to "user-space didn't come up". never mind that the author used the term "almost bricked" to indicated they knew it wasn't hard bricked but intended to communicate that it also wasn't functional. your entire attack stems from you rashly misinterpreting the author's intent here. but even if the author had demonstrated clear technically illiteracy, there's a way for you to add commentary about the definition of bricked without the distraction of attacking the person's technical aptitude. I think everyone knows it's almost impossible to inadvertently hard-brick anything these days

All in all, you, me, and the comment you initially replied to do seem to share the same understanding: to operate linux you need to have a level of tolerance to deal with the occasional issues that come with it. And for many (even technical) people that's not the right balance. Everything else you introduced to the discussion detracted from that point.

I think you might be betraying your own technical literacy gaps with that comment. In many Linux distros, your package manager can update your kernel and boot loader, from userspace. Yes plenty of safeguards exist to try to prevent really bad things from happening. Furthermore, brick has become slang for “doesn’t work”. I dislike the linguistic shift just as much as you but we can give the benefit of the doubt that the point is that everyday Linux uses still encounter massive problems like this with higher than normal frequency.

Which is the whole point, no? There would be more Linux everywhere if it didn’t occasionally break in ways that require special expertise to fix.

Eh, I’d say there is a larger than standard proportion of stalwart know-it-alls here. Doesn’t mean your example is wrong. But man some people have everything figured out and still find a way to complain about it all.

It’s not black and white.

There are companies where you can spend years doing as you describe. More and more, though, you’re competing with people who care even though they don’t own, put in the same effort YoY, and invest in their job. Companies love these employees.

So I guess it really depends on your values and what you want out of life. If you enjoy hobbies and time outside of work, sure find a job where you can coast. Don’t get frustrated when you get laid off just find another place to work. Etc.

Plenty of people want to grow within the industry and build a career, though. And many have what we call basic self-respect and care about how they are perceived.

Without any comparison I’m pretty sure the ones trying to correct LLM code are the same ones that incessantly bike shed human code decisions too.

I think we just have different definitions of execution. Getting VC funding and spending two years honing and SaaSifying an idea then dropping it with a hype train (yes paid for by VC funding) is certainly very different execution than someone writing a paper and publishing model weights on huggingface then vibe coding a “product” around it a year later the day after Jev is announced and the hype proves defensible and the idea is similar.

> At least this post has some touch with the reality that this functionality was available a year ago and was well known among ML people.

When you market a product you make exciting claims relative to the audience you’re engaging with. When was the last time you saw a product marketing page reverently lost all the academic research and prior art that came together to make a product possible?

If Layla’s functionality was available in a SaaS form in a way that could be used by all the people who are excited about and using Jev, wouldn’t this research have won hearts and minds last year when it landed? I would have a lot more empathy for the author if they’d taken a product to market and nobody cared. But even then maybe the market wasn’t ready. There are still reasonable explanations why sometimes ideas take off. We’re on a venture capital forum this shouldn’t need an explanation.

I can understand why the author feels bitter but it still feels juvenile to me. Certainly both Jev and Laya are based on the research of countless prior papers and academics. Diogo decided to build a product out of the concept. The author didn't. Publishing research papers and model weights is probably part of the problem--it feels academic. If you look at the author's profile they focus on applying AI to healthcare. Not selling general AI type safety to AI pilled companies and devs. There's a big difference there. Whether that's good or bad you can argue all day. But for the author to expect otherwise is pretty weird. I do applaud them for not stewing too much on it and trying to do something about it, though.

Time is money. Emotion. You think someone's not going to buy a sports car because it will be cheaper in 10 years? That's 10 years without a sports car.

They might, however, ask whether they need a car with 30% profit margins or one with closer to 10%.

I don't think the grand social goal is to maximize economic growth. It's a stable healthy society.

Right now it really just feels like people spend on whatever the hell they want to without any thought whatsoever. Consume. Yes good, consuuuume. I'd trade slower economic growth for more responsible spending...

I've done my fair share of reversing, modding, rooting, jailbreaking. Asserting my rights is very interesting to me, which is why I think you're missing a fair bit of nuance. I ran this discussion through Astra to check my own recollection and it came to the point that: "copying proprietary firmware and circumventing DRM are separate questions".

The Connectix case doesn't establish a right to circumvent DRM to play games that you "own". It only provides fair use protections for emulator developers.

The Bleem case (the one where the president said the litigation bankrupt them) has to do with game screenshots in comparative advertising.

And probably most apropos, irrespective of whether DRM is used or not, there is no protection in the DMCA or other US code that allows you to circumvent an authentication check in order to play a game. So "just remove the steam dll load command, it's trivial" exposes you to the same level of legal liability as "ugh I have to strip the Sony DRM then crack my copy of <sony exclusive>".

If you look at what's happened recently, there are narrow regulatory exemptions, issued in 2024 and codified at 37 C.F.R. § 201.40(b)(19) that cover, among other things, restoring personal local play of lawfully acquired complete games after required authentication-server support ends. Other provisions permit preservation by eligible libraries, archives, and museums, including limited circumvention of console software. They do not establish a general personal right to jailbreak any console or decrypt any purchased game.

Very respectfully, I think you're conflating a few different things here. I very much wish the world was as you are trying to assert. But my own understanding of US law and case law and a check against Astra don't corroborate your claims.

If there were examples made of legal consequences I think that would at least change the behavior if not solve the problem. Maybe liability should rest with the company that owns the infrastructure running the AI. For most consumer situations that would be the companies developing the models.