The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Is there any non-silly non-delusion argument for why AI is a "scam"?
Poast new message in this thread
Date: June 6th, 2026 12:51 PM Author: The Penis If the goal was just to produce a machine that's more intelligent than humans then the goal has already been met and exceeded. So I don't see any serious intellectual argument for it being flame other than arguments that divert to whether it replaces human workers or not (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918565) |
 |
Date: June 6th, 2026 12:57 PM Author: Smoker It’s not a “scam” but what may be a scam is that it suddenly replaces the whole white collar labor force. Right now it is just enhances productivity. Technically, there’s an argument to be made that it increases the value of labor. No one really knows, but it makes sense if one engineer is suddenly worth ten engineers that you want to hire even more engineers / pay them more. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918579) |
|
Date: June 6th, 2026 1:01 PM Author: The Penis Yeah I agree that it isn't going to replace the labor force. Human institutions exist for reasons other than cognitive labor, and AI is augmentation driven at this point. But AI not causing everyone to stop working is very different from saying it is flame. The main point of AI is to create a thinking machine and that has already been done so it isn't flame. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918593) |
|
Date: June 6th, 2026 3:45 PM Author: .,..,,.,.,.,.,.,..,.,.,..,.., CR. The idea it will make a human workforce unnecessary is kind of like the way people briefly said we’d all be riding Segways (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918854) |
|
Date: June 6th, 2026 1:04 PM Author: Microsoft Execution Containers It's been replacing some low level shitjerbs for a while now, but nothing highly skilled. Anyone who got replaced by AI can find a jerb at Home Depot or UPS (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918601) |
|
Date: June 6th, 2026 1:11 PM Author: The Penis Yeah that still makes it sound like AI has only replaced dumb people though and will some day maybe replace everyone else. But in reality AI is already more intelligent than the smartest surgeons at problem solving and reasoning by a wide margin. So it is really just about whether what you do is coupled to real world physical consequences or social systems. But again AI doesn't even have to replace humans to be legit anyway.
Deep sea saturation divers, weapons test engineers, venomous animal handlers etc are also now >>>>>>>>>>>>>>> more prestigious than lawyers, professors and software engineers until we build actual physical robots that are intelligent. |
Date: June 6th, 2026 1:01 PM Author: .,.,.;;,;.,;:,:,,:,.,:,::,..;.,:,.:;.:.,;.:.,:.::, everyone who makes this argument does not come from a world in which its material benefits are obvious and measurable, all they see is a machine that can generate writing and videos, and they are very happy to conclude from that that it is all slop. which is true from what they see of it but is also a relieving thought that they're happy not to interrogate any further.
if they had any serious experience coding or doing drug discovery or anything similar they'd understand that this question of "being as smart as a human" or "moving beyond slop" is not the point, it's enough for it to be a several-orders-of-mangitude step up in processing power and an opening up of the ways in which humans can tap into it with natural language. |
|
Date: June 6th, 2026 1:05 PM Author: The Penis A lot of the slop is from older models too or is artwork. The stuff it produces can be clunky aesthetically for a human but doesn't mean it isn't powerful as a reasoning engine. It solves problems humans could never solve practically on it's own now and it augments humans in ways that massively increase throughput. That's enough. Replacing jobs is a different thing, because you can use it to just do more. Although it will obviously displace some people when people realize it's better to have one senior with AI instead of hiring a bunch of juniors (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918603) |
|
Date: June 6th, 2026 1:26 PM Author: cowgod63&1/3 i believe ai has awe-inspiring potential and will eventually completely overhaul the human labor market, perhaps ultimately in a manner more sweeping than the industrial revolution. that said, people are still often wont to proclaim it a “scam” because loudmouth midwits like op are frantically running around for no discernible reason saying stupid things like it is already “orders of magnitude more intelligent” than humans. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918610) |
|
Date: June 6th, 2026 1:45 PM Author: The Penis This is a low iq take. Orders of magnitude more intelligent than humans is already measurable and arguable if you declare what metrics you are using to define intelligence. Eventually more sweeping than the industrial revolution in terms of labor market dynamics is just speculation. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918627) |
|
Date: June 6th, 2026 1:52 PM Author: cowgod63&1/3 “actually, my argument is correct if you unquestioningly accept my retarded postulates.” (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918635) |
|
Date: June 6th, 2026 2:00 PM Author: The Penis I'm not saying you can't question them. But come on. When you build a system that doesn't have the cognitive bottlenecks of humans and has absorbed millions of books, has essentially instantaneous retrieval, no forgetting from interference, no fatigue, and the ability to operate across hundreds of domains. Plus this thing is finding proofs Erdos missed, finding counterexamples Erdos missed, connecting fields Erdos never connected, and doing so while simultaneously being competent in thousands of other domains. Erdos was already at the extreme tail of the human population. When you take these things into account does orders of magnitude really sound so crazy? We always measured human IQ with tests like the old gre and old lsat (before they got nerfed) and AI has been able to max those for several generations now. When AI dominates these tests, it is special pleading to suddenly declare that those measures no longer count merely because the performer is nonhuman. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918671) |
|
Date: June 6th, 2026 2:07 PM Author: cowgod63&1/3 current ai models are not (at least yet) independent agents. they are tools. is a casio calculator watch “more intelligent” than a human because of its mathematical processing power? is a microsoft encarta cd-rom “more intelligent” than a human because of its breadth of “knowledge” and perfect recall? the answer to both of those questions is not “no” because the questions aren’t even validly formed in the first place. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918712) |
|
Date: June 6th, 2026 2:16 PM Author: The Penis That analogy is just bad. A casio calculator is superhuman at arithmetic, but it isn't doing general reasoning. It doesn't parse a novel natural language problem, decide which abstraction applies, test solution paths, explain the result etc. Encarta has stored information, but AI is not merely retrieving info from a database, it is drawing inference and integrating across a compressed statistical representation of a massive corpus and it can answer novel questions by doing it. Calculators and Encyclopedias requires humans to formulate the problem 100% of the way. But with frontier AI humans are outsourcing parts of the cognitive process that were always traditionally considered Intellectual Work.
Tool not an agent also doesn't refute cognitive ability. Agency and intelligence are separable. You don't need autonomy, persistent goals, phenomenology or any of that nonsense to have reasoning, abstraction and problem solving ability. You are just goalpost moving to things that fall outside of what has traditionally been defined as "intelligence". Calculators and encyclopedias require humans to formulate the problem 100%. With frontier AI humans are outsourcing parts of the process that were always traditionally considered Intellectual Work. |
|
Date: June 6th, 2026 2:51 PM Author: cowgod63&1/3 “But with [abaci/the difference engine/eniac/microprocessors/etc.] humans are outsourcing parts of the cognitive process that were always traditionally considered Intellectual Work.” there’s nothing particularly novel about the thrust of that claim. if ai really is so smart without the need for human input or direction, why don’t you go use it to scoop up the $6 million remaining dollars for the millennium prize problems? |
|
Date: June 6th, 2026 2:59 PM Author: The Penis That's just moving goalposts again. I never claimed it could instantly solve every famous open problem. We don't even know that all Millennium problems are solvable in the relevant sense, could require new foundations, decades of theory building etc. Von Neumann was orders of magnitude smarter than the average human and he didn't solve the Riemann hypothesis. I never claimed it was omnipotent across all hard problems. It's superhuman on many metrics used to measure intelligence and has solved a few open problems already that genius humans never solved themselves. That's already pretty impressive imo. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918829) |
|
Date: June 6th, 2026 2:22 PM Author: TurboGrafx-67 wtf are you smoking? Google agentic orchestration and then go outside and touch the ground (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918767) |
|
Date: June 6th, 2026 1:32 PM Author: cowgod You need to separate your issues imo
Highly powerful tool circa 2026: yes
Future state (not 2026, maybe 2030s): perhaps unimaginably powerful if it can Engineer things insanely quickly: yes
Insane resource hog: yes. Which kind of defeats the purpose or limits the upside imho
A Scam: maybe, because literally everything “big tech” norcal SV Nerds have ever done for the last 25 years has been either a failure (google glass, metaverse) or not value accretive. Or it just Destroyed newspapers in a way that adds no real value. They transferred value to themselves rather than creating new value. I don’t think “tech companies” have ever “developed technology” afaik
|
|
Date: June 6th, 2026 1:46 PM Author: The Penis "Scam" is besides the point when it has already been built and it works. Of course everything is a scam to some extent in that people are constantly trying to sell things to other people they don't need or want. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918628) |
|
Date: June 6th, 2026 1:55 PM Author: cowgod The point imho is that with AI they’re asking you to take a leap of faith and assume that it continues to rapidly get better. And that’s a lot to ask when big tech literally fails at everything. They’ve never “done anything” other than disrupt existing industries and make themselves richer. They Fail Constantly at doing actual things while growing market cap from algorithms wreaking havoc on Proles’ brains.
The list goes on and on and on. Self-driving cars were being talked about the same way 10 years ago. It’s the same basic idea. A technology that’s technically better than humans on average in many ways but which needs Oversight. Google Glass Google+ Google Wave Google Buzz Google Allo Google Duo Google Stadia Google Reader Google Hangouts Google Talk Google Notebook Google Video Google Answers Google Health Google Lively Google Daydream Google Spaces Google Latitude Google Code Google Toolbar Google Currents Google Podcasts Google Jamboard Google Clips Project Ara Project Loon Google Fiber (national ambitions) Waymo (original timelines) Google Wallet (first version) Google TV (first version) Metaverse Horizon Worlds Facebook Home Facebook Portal Facebook Dating Libra Diem Novi Workplace Facebook Deals Facebook Credits Facebook Camera Rooms Poke Slingshot Paper Microsoft Windows Phone Zune Kinect Mixer Cortana Internet Explorer Mobile Microsoft Band Windows RT Groove Music Kin MSN Music HoloLens Bob Zune
If this tech were being driven by companies that Do Things and aren’t Jewish like Mitsubishi Heavy Industries and Siemens + Konami/Capcom/Squaresoft, people would be less skeptical. “Big tech” are just Evil Nerds who add no value to anything in 25 years. Guilt by association. |
|
Date: June 6th, 2026 2:02 PM Author: The Penis Yeah but the difference is that AI is already big at its current development point without it getting any better. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918680) |
|
Date: June 6th, 2026 2:04 PM Author: cowgod Imho it’s Almost Big. To be big, it needs to be Turn-Key. There is no Turn Key AI product. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918687) |
|
Date: June 6th, 2026 2:08 PM Author: The Penis I think it's only a matter of time now until it reaches Turn-Key status and receives the blessing of the cowgod. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918717) |
|
Date: June 6th, 2026 1:46 PM Author: Genius Bear on the loose in Japan It's a scam in the sense that it's a deliberate concentration of power by self-interested agents, to the detriment of the members of the population who are not within that coterie of self-interested agents who benefit from that artificial concentration of power (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918629) |
|
Date: June 6th, 2026 1:48 PM Author: The Penis I agree but my point is that that doesn't really matter when we already have thinking machines that are smarter than humans. That is big enough in itself to make it look silly when people try to say it's flame. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918631) |
|
Date: June 6th, 2026 1:53 PM Author: Genius Bear on the loose in Japan You're talking about something completely different which doesn't fall under the criticism of "scam" and would require different language and definitions in order to narrow the topic to something legible (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918636) |
|
Date: June 6th, 2026 2:03 PM Author: The Penis Well everything is. $cam. My point is that it isn't flame in the sense that the tech is real. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918682) |
|
Date: June 6th, 2026 2:05 PM Author: Genius Bear on the loose in Japan Should have used the word "flame" instead of "scam" then, Champ (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918697) |
|
Date: June 6th, 2026 2:06 PM Author: The Penis That's fair. I have no problem admitting tech companies are $cammy in how they market and hype products. But I stand by my statement that AI is Not Flame (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918707) |
|
Date: June 6th, 2026 1:55 PM Author: .,.,.;;,;.,;:,:,,:,.,:,::,..;.,:,.:;.:.,;.:.,:.::, in other words it's a "scam" in the sense that you want to assign words with negative connotations to shit you don't like. fucken brilliant mate. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918648) |
|
Date: June 6th, 2026 2:01 PM Author: Genius Bear on the loose in Japan The accurate version of this response would be that I want to assign words with negative connotations to things that harm me and my interests (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918672) |
|
Date: June 6th, 2026 2:09 PM Author: TurboGrafx-67 It's only a "scam" in that the compute power to drive current AI usage isn't sustainable at cost but the elite/institutional class and top 1% are ok with burning all that cash for now because over 99.999% of the benefits of AI will flow upstream to their pockets. But this isn't breaking news insofar as elections, taxes, stock market, popular media etc are all part of the same huge scam (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918720) |
|
Date: June 6th, 2026 2:29 PM Author: tokens in abundance I'm sure between all the datacenters, new power plants, Nvidia getting more compute power per megawatt, the whole limiting factor being electricity thing will sort itself out. Running out of electricity I don't think is the bottleneck. This is just cope. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918790) |
|
Date: June 6th, 2026 2:31 PM Author: TurboGrafx-67 Yes it will be sorted out. Top 1% gets best in class frontier AI with no guardrails. Everyone else gets watered down instagram level shit (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918798) |
|
Date: June 6th, 2026 2:35 PM Author: tokens in abundance It's not just top 1% it's going to be Mag7 businesses and all the FANG companies and everyone trying to compete. That's how business works. They don't give cloud computing and Oracle tools to kids either. That's what business is, pay to play. Why should Timmy Dipshit get frontier AI for free? Life costs money that's how everything works. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918804) |
|
Date: June 6th, 2026 3:06 PM Author: TurboGrafx-67 You aren't disagreeing me and nothing I said was prescriptive
The difference between "just how business runs in 2026" and "scam" isn't some bright hard line |
|
Date: June 6th, 2026 3:15 PM Author: tokens in abundance Your Tone is disturbing. You don't sound like a Shareholder. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918835) |
|
Date: June 6th, 2026 2:27 PM Author: tokens in abundance Being able to replace low level white collar workers is not a scam. It's an innovation. If Toyota found a way to make the Camry for 30% less it would be an innovation.
You're right about how proles are detached from the end product. I don't believe AGI is coming soon. It's not going to write new plays that are as good as Shakespeare. It's not that creative. But it is going to make businesses, every business, but especially tech businesses, way more efficient. It's going to be cheaper and easier to do anything and everything, rapidly.
As such the real innovations are the ones that haven't happened yet because some kid in San Francisco hasn't come up with it yet. People thought the internet was fucking incredible in the 80s and 90s and think of how far they were from Amazon, Tesla, and DoorDash. The platform is not the end product. AI is a platform. What will amaze the general public are the innovations yet to come. Some small group of engineers will have an idea that changes the world and AI will make it possible.
Look at how many billion dollar companies no one could have foreseen in 2010. Multiple bitcoin trading firms. Polymarket and Kalshi. Rockets that can land and be reused. Self driving cars. This is just the shit that not even super smart people could really visualize in 2010. AI is going to create a revolution where humans in conjunction with machines lead to massive innovations and new technologies cropping up everywhere on a grand scale. But yeah I don't think it will be able to write a book as good as War and Peace. That's what proles are saying when they say it's a scam. It didn't make a feature length film as good as The Godfather therefore it's shit.
No but it's obviously going to turbocharge technology companies send everything into overdrive. |
|
Date: June 6th, 2026 2:48 PM Author: the walter white of this generation (walt jr.) The amount that's being invested in it means that it needs to hit crazy benchmarks of real-world impact. Right now, it's not an invention on the order even of the internet; in fact it's mostly being used to synthesize information that's already on the internet.
I would very much like to see a bullish-case AI future. I'll be among the last replaced in my field and will reap the benefits of labor replacement in the meantime.
But I think maybe the chinese model of emphasizing last-mile integration and IRL use cases over raw power might be superior. I'd be fine with the fact that it still can't write a good brief and hallucinates every 4th case it gives me, if it could handle rote legal-assistant work like downloading all my case filings as they come in on CM/ECF and my shitty state court system, converting them to PDFs and naming them in my preferred convention, putting them in my case file, and calendaring any deadlines that are triggered by them. But it can't do that; I've both tried to make ChatGPT do it and tried a free trial of a program that purports to be able to do it, and they've been worse than the most retarded of legal assistants. And I can't tolerate a 95% accuracy rate on calendaring hearings. |
|
Date: June 6th, 2026 2:57 PM Author: cowgod Shut up queer (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918827) |
|
Date: June 6th, 2026 3:02 PM Author: tokens in abundance The internet had decades of run up before becoming the internet as we know it today. AI is similarly in its infancy which drives early adopter mindset. (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918831) |
|
Date: June 6th, 2026 3:13 PM Author: TurboGrafx-67 It can do complex lexis research and intelligently extract insights from shepards and judicial analytics from lex machina and use all of that to write briefs without hallucinations or em dashes.
Your experience does not reflect the actual capabilities as we sit here right now. I don't know why and not interested in getting into the weeds on your practice. Suffice to say you aren't close to maxing out current usage. |
|
Date: June 6th, 2026 3:15 PM Author: TurboGrafx-67 Also just quickly there a ton of pdf solvers on GitHub to optimize pdf for cloud AI or convert to markdown if more optimal. I just used Claude to code my own
You can also subscribe to a platform like docketbird and integrate it with gpt/claude and Google Calendar |
|
Date: June 6th, 2026 3:47 PM Author: .,..,,.,.,.,.,.,..,.,.,..,.., Not all AI is a scam but plenty of scammers love to leverage the idea of AI in their scams (http://www.autoadmit.com/thread.php?thread_id=5871621&forum_id=2Vannesa#49918856) |