Invading Mexico to End the Cartels
- Thread starter Birdjaguar
- Start date
Bestbank Tiger
Deity
It's a figure of speech. Just means somebody is putting too much faith in something.
Hygro
soundcloud.com/hygro/
I’m going to cut to the chase, imaginative people have been believing the tech industry has far greater data collection and sorting abilities than it does or will, for a long time now. AI that can do what you imagine hasn’t been figured out yet. We don’t have the hardware. It’s a monumental task and everyone’s taking a big paycheck for much less work.
The conceptual way they could be doing it at a high level is trivial, which is why I used to think they were doing it already, years ago, too. But in engineering, big picture thinking is not the bottleneck.
Hygro
soundcloud.com/hygro/
Sure, by my late teens as well, with prohibition being the obvious example. It didn’t take long for logic and history to show the way. But the reaction to the perceived threat of drugs by so many is visceral.
Joij21
🔥Hail Satan!🔥
Sounds like a cover up to conceal true capabilities.
Look they were saying they were already capable of this sh** AND DOING IT over ten years ago! (Major respected media outlets, and leakers like Edward Snowden who worked in the upper echelons of all this data gathering himself).
Saying they don't have the capability now when everyone said they did a decade ago is essentially insinuating that somehow they lost the technical ability and big tech must have undergone some kind of dark age or something in the years between then and now.
The only things I can think of that could lead to such a "little dark age" would be the people who did have the skills to know how to operate these systems are no longer there (not enough pay, pay not keeping up with inflation, retired, fired, etc.), subsequent software got too complex whereby the new guys can't understand it or it's too buggy and was never fixed since the old guys who were working on it are now gone, and the new guys willing to work for the current pay that silicon valley is willing to fork out just aren't motivated nor as skilled coming out of the universities (with the few creme da la creme ones demanding such higher pay to even consider work, that they are just outright rejected by the valley's companies leading to an overall brain drain), and with poorer skilled workers they can't perfect the AI models that that the earlier more experienced generation of programers set up before abandoning halfway and leaving to greener pastures thus forcing them to rely on open source methods of AI training that will make such AI extremely difficult to commodify going forward (and therefore will eventually stagnate at some point to where it will never surpass human capabilities simply because there would be no profit in making better AI once everyone realizes no one can quite easily get an IP monopoly on it).
In summary Silicon Valley is currently experiencing a dark age isn't it? (I mean @Hygro you used to work in tech so maybe you know)
Joij21
🔥Hail Satan!🔥
Hygro
soundcloud.com/hygro/
The NSA was tasked by Congress to monitor the entire internet so they “did”. They were overwhelmed with data and were happy to shut it down and go back to something more manageable when there was backlash. They don’t have genius hackers coming up with creative solutions to every problem, they have enterprise software and a bureaucracy and a mandate to track certain groups which in the time of their “monitor the whole internet” were religious terrorists and 2nd world hackers. They were busy with their job.
Data is expensive. It wasn’t permanently stored, they don’t have AI that just knows how to sort it. The new useful AI today doesn’t sort it, it takes it to generate new data which is useless for this purpose.
The difference between being targeted and known and being automatically known when caught in a dragnet of data is everything. If the FBI wants to know you, subpoenas your ISP and gets access to your accounts, they’ll know you. They will need at least an agent reviewing your information. The FBI is a little busy and took quite a while tracking down people whose faces were in videos in the capitol. This is not an unlimited org. Neither is the NSA with a backlog of military and espionage concerns.
Facebook and Google have the best AI driven revenue technology. It uses targeted advertising. Google had my profile very wrong ten years ago. Facebook, which has direct info, finally figured out things to sell me in 2022 for a little while. Mostly ads for things I owned or adjacent products I would never want, but it fed me sales on a few products I wanted. Made sense I was buying a ton of those products on the browser profile with my Facebook. Since not buying more, they no longer can figure out what I buy. They advertise a lot in Spanish, which is cute but backward. AI at its best can anticipate genre but never quality.
A great example of that is Spotify which has one of the best recommendation algorithms. It’s a low bar. If you don’t work with its confines, it gives you terrible recommendations. You have to save way too many songs to your library before it even gets close to knowing, and even then, it has no idea you might be discerning on axes of quality, which as far as it is algorithmically concerned, is a black box of an arbitrary subjective mess. And it is, except that it totally isn’t. Believe me I have talked to gpt4 about it, the existing literature cannot synthesize an understanding of that quasi objective quality that people “get” when they are awake to it. But Spotify also completely fails to deep learn or spontaneously match people on a “I only want the best in xyz genres” with the others like that. No, it is a highly parameterized monstrosity ultimately stuck on genre, limited to the creative efforts of its team, mostly crystallized some years back.
Reddit has an incredible sophisticated recommendation system. It too has to same fundamental design flaw and will eventually just give you whatever thing you looked at one time over and over again to the detriment of things you actually and regularly want to read. Often times a fresh browser Reddit gives me better content for me than what I log in and see.
These companies, and especially google and Facebook, have funded much of the AI research and have the clearest channels and biggest incentives to use it, and they are mid at best. They’ve had me on the same accounts for almost 20 years and Facebook alone could barely advertise to me for a few good months.
There’s a bench of a girls memorial up the hill from me. She died I think 2011, born maybe 1988. Everything was online by 2011, especially deaths and formal things like county benches. She doesn’t even exist. Some people aggregators have her “living” at a couple different last addresses, no information. No obituary and for sure there was one, her family got her a stoner bench in a choice spot ffs! Not online.
AI and the hardware has never been better, and it’s all still trash except for tools here and there like using chat gpt to help you with a topic that is well documented like coding. Still requires a coder to make it work though, it’s laden with hallucinations. The difference between putting agents on investigating a person and getting meaningful mass data is a world apart, and has not been solved. Storage is expensive and while a few companies are trying to save what they can, most are trying to dump that information as fast as possible.
Data is expensive. It wasn’t permanently stored, they don’t have AI that just knows how to sort it. The new useful AI today doesn’t sort it, it takes it to generate new data which is useless for this purpose.
The difference between being targeted and known and being automatically known when caught in a dragnet of data is everything. If the FBI wants to know you, subpoenas your ISP and gets access to your accounts, they’ll know you. They will need at least an agent reviewing your information. The FBI is a little busy and took quite a while tracking down people whose faces were in videos in the capitol. This is not an unlimited org. Neither is the NSA with a backlog of military and espionage concerns.
Facebook and Google have the best AI driven revenue technology. It uses targeted advertising. Google had my profile very wrong ten years ago. Facebook, which has direct info, finally figured out things to sell me in 2022 for a little while. Mostly ads for things I owned or adjacent products I would never want, but it fed me sales on a few products I wanted. Made sense I was buying a ton of those products on the browser profile with my Facebook. Since not buying more, they no longer can figure out what I buy. They advertise a lot in Spanish, which is cute but backward. AI at its best can anticipate genre but never quality.
A great example of that is Spotify which has one of the best recommendation algorithms. It’s a low bar. If you don’t work with its confines, it gives you terrible recommendations. You have to save way too many songs to your library before it even gets close to knowing, and even then, it has no idea you might be discerning on axes of quality, which as far as it is algorithmically concerned, is a black box of an arbitrary subjective mess. And it is, except that it totally isn’t. Believe me I have talked to gpt4 about it, the existing literature cannot synthesize an understanding of that quasi objective quality that people “get” when they are awake to it. But Spotify also completely fails to deep learn or spontaneously match people on a “I only want the best in xyz genres” with the others like that. No, it is a highly parameterized monstrosity ultimately stuck on genre, limited to the creative efforts of its team, mostly crystallized some years back.
Reddit has an incredible sophisticated recommendation system. It too has to same fundamental design flaw and will eventually just give you whatever thing you looked at one time over and over again to the detriment of things you actually and regularly want to read. Often times a fresh browser Reddit gives me better content for me than what I log in and see.
These companies, and especially google and Facebook, have funded much of the AI research and have the clearest channels and biggest incentives to use it, and they are mid at best. They’ve had me on the same accounts for almost 20 years and Facebook alone could barely advertise to me for a few good months.
There’s a bench of a girls memorial up the hill from me. She died I think 2011, born maybe 1988. Everything was online by 2011, especially deaths and formal things like county benches. She doesn’t even exist. Some people aggregators have her “living” at a couple different last addresses, no information. No obituary and for sure there was one, her family got her a stoner bench in a choice spot ffs! Not online.
AI and the hardware has never been better, and it’s all still trash except for tools here and there like using chat gpt to help you with a topic that is well documented like coding. Still requires a coder to make it work though, it’s laden with hallucinations. The difference between putting agents on investigating a person and getting meaningful mass data is a world apart, and has not been solved. Storage is expensive and while a few companies are trying to save what they can, most are trying to dump that information as fast as possible.
Narz
keeping it real
The surveillance state is concerning but on the other hand 99.99% of humans just aren't interesting enough to warrant much attention from it.
On the other hand still be careful what you say in case you ever do become famous.
Or in case your afterlife is based on the quality of your online fingerprint #pascalswager #justincase
On the other hand still be careful what you say in case you ever do become famous.
Or in case your afterlife is based on the quality of your online fingerprint #pascalswager #justincase
I have not read the rest of the thread, but the answer here has to be drug legalisation. If the US sells guns freely then Mexico should sell drugs freely. Set up shops on the border selling high grade government regulated drugs and the cartels would be out of business within Mexico in a day.
