User Panel
User Panel
User Panel
|
Posted: 2/2/2023 2:58:05 PM EDT
Given that computer science happens to be my area of expertise, I generally find the fear of AI, our "robot overlords," and all of that to be comically overblown, and the discussions usually end up reminding me of my half-drunk and half-retarded uncles talking about the Nazca lines or some other dumb shit they saw on the History Channel. There will never be a "singularity" and there will always be things that an AI fundamentally cannot do. They won't replace people.
But what they CAN do is fundamentally alter reality. Speech synthesis and video editing are not new. But what AI/ML speech synthesis and image generation models, like Unreal PaGAN, enable is the almost real-time creation of audio and video that are not real. Go search the DeepTomCruise stuff for a simple example. It's not the wonder of the "this person doesn't exist" websites. It's the recognition that even high-quality video interviews can be produced of real people, easily able to convince those who know them best, that never happened. There are evil people, and they would be the ones who would exploit these things. Imagine not being able to believe video and audio of a person you knew.  Pinscreen PaGAN: AI-Generated Real-Time Avatars in UE4 |
|
|
|
Posted: 2/2/2023 3:03:03 PM EDT
[#1]
The difficulty in figuring out what is real and what is fake will be effectively impossible if you didn't witness it yourself. Confirmation bias ftw
|
|
|
|
Posted: 2/2/2023 3:06:09 PM EDT
[#2]
Just out of curiosity, why do you believe that there will never be a singularity event?
A lot of smart people believe that there will. Why do you believe that there won't? |
|
|
|
Posted: 2/2/2023 3:15:51 PM EDT
[#3]
Quote HistoryQuoted: Just out of curiosity, why do you believe that there will never be a singularity event? A lot of smart people believe that there will. Why do you believe that there won't? A lot of smart people also believe that it won't. Futurists are notoriously fantastical and optimistic. But human intelligence is multi-dimensional, and computer intelligence, however advanced it may become, will never be the same. It will never be capable of creativity, or emotion, or even a general ability to understand "why." Part is this is because they can only ever do what they're told, and they have no fundamental ability to understand if something "learned" is right or wrong. Either it's part of their training set or it's not. Unsupervised machine learning is like a fever dream on LSD because "artificial intelligence" without reinforcement is almost meaningless. |
|
|
|
Posted: 2/2/2023 3:19:27 PM EDT
[#4]
OP is a bot.
|
|
|
|
Posted: 2/2/2023 3:20:34 PM EDT
[#5]
They're coming for you.
 Robotic dog armed with submachine gun |
|
|
|
Posted: 2/2/2023 3:22:03 PM EDT
[#6]
this reminds me of 2014 when i was managing a team of 20 or so CCNP level network guys around the world and they were trying to convince me it was unpossible to automate networking or do software defined networking. lol
|
|
|
|
Posted: 2/2/2023 3:25:02 PM EDT
[#7]
OP works for Cyberdyne Systems
|
|
|
|
Posted: 2/2/2023 3:25:08 PM EDT
[#8]
Quoted: Given that computer science happens to be my area of expertise, I generally find the fear of AI, our "robot overlords," and all of that to be comically overblown, and the discussions usually end up reminding me of my half-drunk and half-retarded uncles talking about the Nazca lines or some other dumb shit they saw on the History Channel. There will never be a "singularity" and there will always be things that an AI fundamentally cannot do. They won't replace people. But what they CAN do is fundamentally alter reality. Speech synthesis and video editing are not new. But what AI/ML speech synthesis and image generation models, like Unreal PaGAN, enable is the almost real-time creation of audio and video that are not real. Go search the DeepTomCruise stuff for a simple example. It's not the wonder of the "this person doesn't exist" websites. It's the recognition that even high-quality video interviews can be produced of real people, easily able to convince those who know them best, that never happened. There are evil people, and they would be the ones who would exploit these things. Imagine not being able to believe video and audio of a person you knew.  https://www.youtube.com/watch?v=7YlJ28r1-J4I don’t see that being a problem. Even since there were humans, there have been fakes. We survived over a million years. The real concern is the disruption that AI will cause. You see the artists already complain about AI generated content and graphics. Wait until millions are out of jobs. |
|
|
|
Posted: 2/2/2023 3:26:26 PM EDT
[#9]
that looks fake tho
|
|
|
|
Posted: 2/2/2023 3:30:44 PM EDT
[#10]
AI is and will continue to be used to create (slanted, biased, agendized) web content, ultimately solving the "problem" (to the oligarchs, WEF, lizard people, whatever) of the internet still (somewhat) being a medium of open expression and dissemination of actual, objective fact. This content is already favored by major (think Google) search engine indexing.
TL;DR - AI will add a LOT of noise to the signal when it comes any web-accessible material that strays from the narrative. |
|
|
|
Posted: 2/2/2023 3:36:32 PM EDT
[#11]
I’m pretty sure this is already what happened to Hannity, and other formerly respected conservatives.
It’s the only thing that makes sense. Liberals, diligently making “conservatives” look as stupid and narcissistic as possible. |
|
|
|
Posted: 2/2/2023 3:43:26 PM EDT
[#12]
We are there. And beyond.
 Sue Saad - Looker Sorry, let me put in a video that doesn't have so much T&A. RvX1FvO6oJ0 for those interested. |
|
|
|
Posted: 2/2/2023 3:52:33 PM EDT
[#13]
Quote HistoryQuoted: I don’t see that being a problem. Even since there were humans, there have been fakes. We survived over a million years. The real concern is the disruption that AI will cause. You see the artists already complain about AI generated content and graphics. Wait until millions are out of jobs. Of course fakery has been around forever. I think the problem is that AI/ML can generate a convincing audio/video clone of a real person in a matter of minutes, and it can do so under the control of a malicious actor. |
|
|
|
Posted: 2/2/2023 3:54:08 PM EDT
[#14]
Quote HistoryQuoted: A lot of smart people also believe that it won't. Futurists are notoriously fantastical and optimistic. But human intelligence is multi-dimensional, and computer intelligence, however advanced it may become, will never be the same. It will never be capable of creativity, or emotion, or even a general ability to understand "why." Part is this is because they can only ever do what they're told, and they have no fundamental ability to understand if something "learned" is right or wrong. Either it's part of their training set or it's not. Unsupervised machine learning is like a fever dream on LSD because "artificial intelligence" without reinforcement is almost meaningless. I'd say that's probably accurate regarding current paradigms, but l'm wary of that word "never". If we're talking the timescales that futurists have typically predicted for it, I'd say you're probably right. I don't think that we'll see it any time soon. But what is this shit gonna look like in 20 years? 50 years? I don't think that anyone can honestly say, aside from "it's gonna be different". I'm more concerned about the near term. AI tools are going to be used by evil people to perform a "soft" enslavement of the populace. It's already happening. We're increasingly living in a panopticon, everything that we do is tracked, and evil people are trying to push us into technologies that'll enhance their ability to control our behavior. Very soon we will find ourselves in a situation where they can "non-person" you for misbehaving. They'll be able to shut the power off to your home, deny access to your bank accounts, restrict your ability to purchase things, and track every single thing that you do, and every place that you go. What leads me to think that a singularity might be in the cards at some point is that it's clearly become an arms race, and people with less scruples than your average programmer will inevitably be involved. If there's a way for it to happen, then in this case it seems inevitable. |
|
|
|
Posted: 2/2/2023 3:58:02 PM EDT
[#15]
Quote HistoryQuoted: this reminds me of 2014 when i was managing a team of 20 or so CCNP level network guys around the world and they were trying to convince me it was unpossible to automate networking or do software defined networking. lol SDWAN, Cloud this cloud that etc. etc. Still in many cases need a ton of manual intervention. |
|
|
|
Posted: 2/2/2023 4:01:10 PM EDT
[#16]
How long before nuke/EMP is employed by a rogue state, has-been, or never-was entity that is tired of being pushed around by the machine?
How durable would an AI hegemony be through a coronal mass ejection or similar event? Even if it happens, unless it can get off this planet it's ultimately doomed. Not that that makes me feel any better about the possibility of it happening in my or my kids' lifetimes. |
|
|
|
Posted: 2/2/2023 4:05:11 PM EDT
[#17]
I think this will eventually drive people away from technology. If you can’t believe what you see, hear and read on devices we’ll be right back to face to face interactions.
|
|
|
|
Posted: 2/2/2023 4:09:27 PM EDT
[#18]
Quote HistoryQuoted: The difficulty in figuring out what is real and what is fake will be effectively impossible if you didn't witness it yourself. Confirmation bias ftw It already is. The rubicon has been passed. And if you think it isn't used to move public opinion even today, you are fooling yourself. |
|
|
|
Posted: 2/2/2023 4:27:51 PM EDT
[#19]
Quote HistoryQuoted: Of course fakery has been around forever. I think the problem is that AI/ML can generate a convincing audio/video clone of a real person in a matter of minutes, and it can do so under the control of a malicious actor. Every piece of news today is fake. Our history books are fake. Not much is real nowadays. |
|
|
|
Posted: 2/2/2023 4:40:06 PM EDT
[#20]
 
|
|
|
|
Posted: 2/2/2023 4:46:04 PM EDT
[#21]
Quote HistoryQuoted: I don’t see that being a problem. Even since there were humans, there have been fakes. We survived over a million years. The real concern is the disruption that AI will cause. You see the artists already complain about AI generated content and graphics. Wait until FIFY |
|
|
|
Posted: 2/2/2023 4:50:29 PM EDT
[#22]
This is going to require any/all video and photos that have any legal standing to have encrypted and embedded end-to-end digital signatures based in a distributed blockchain that's big enough and distributed enough to be immune to a 51% attack.
And the architecture is going to require it can handle millions of signatures, transactions, and verifications per second. And various social media platforms may not allow upload or sharing of images/videos that aren't signed. There's going to be a big push for this within a few years. And depending on how the system is designed and operated, it could be impossible to take a picture or a video anonymously. That's cute, but it misses a big point, that while machine learning and weak-AI may well have exponential growth in capabilities, especially if/when machine learning and weak-AI is turned onto tasks directed at improving itself, there's no reason to believe that this will create "real AI", or strong-AGI. A system that has true self-awareness and aspects of metacognition and actual executive agency. And just making them "bigger" is no guarantee of getting there either. The most powerful systems we have are still fundamentally as "dead" or "flat" as a copy of MS-Word. They process data and have zero awareness of what the data is, what it's for, or that the system itself exists. The good news is that the potential for weak-AI/ML to produce ever more complex and life-like results means there's possibly little economic or competitive incentives to strive for strong-AGI. It may actually be a benefit to the most scary applications that involve misinformation, social manipulation, weapons, destruction, and killing. Instead of running amok, a self-aware system could possibly refuse to carry out its mission. Bad news, a human, or humans, with all our flaws and faults, will be in control of such systems. More bad news, even if weak-AI and ML is not on some exponential path to a critical-mass that produces strong-AGI, and there's less and less benefits to making one, people will try anyway. Just for the hell of it. For curiosity. Or simply to see if they can. But at least to date, nothing has approached the self-awareness or independent action of even a small animal. |
|
|
|
Posted: 2/2/2023 4:52:14 PM EDT
[#23]
Automation tied into utility grids, networked air and ground traffic control, generative "deepfake" photos/videos, social media bots, generated articles and reports…The "scary punchline" is that we're rapidly approaching a point where someone could fabricate a major event (or a campaign of minor ones) in a manner convincing enough as to trigger global responses. We may already be there, and it's just a matter of time before some 4channer fakes a nuke convincingly-enough to start armageddon. The consumer-grade image-generating AI people lose their minds over is a baby toy in comparison to what's cutting-edge even just in the commercial sector.
|
|
|
|
Posted: 2/2/2023 4:55:44 PM EDT
[#24]
I fear the people that want to control the AI more than the AI itself.
|
|
|
|
Posted: 2/2/2023 8:22:27 PM EDT
[#25]
Quote HistoryQuoted: Automation tied into utility grids, networked air and ground traffic control, generative "deepfake" photos/videos, social media bots, generated articles and reports…The "scary punchline" is that we're rapidly approaching a point where someone could fabricate a major event (or a campaign of minor ones) in a manner convincing enough as to trigger global responses. We may already be there, and it's just a matter of time before some 4channer fakes a nuke convincingly-enough to start armageddon. The consumer-grade image-generating AI people lose their minds over is a baby toy in comparison to what's cutting-edge even just in the commercial sector. You mean a lie like Trump is a Russian asset? Or like global warming? Recycling? Communism? |
|
|
|
Posted: 2/2/2023 8:27:01 PM EDT
[#26]
I'm not important enough to be deepfake porned.
|
|
|
|
Posted: 2/2/2023 8:28:12 PM EDT
[#27]
Quote HistoryQuoted: FIFY Thanks! You are correct. Last time when we had this kind of creative disruption (the first and second Industrial Revolution), we got two world wars and communism. Just think about when billions of people are out of job and can’t pay their bills. Today the AI revolution won’t last decades - ChatGPT had millions of users in 5 days! This time, the revolution will spread in days and months. |
|
|
|
Posted: 2/2/2023 8:41:41 PM EDT
[#28]
Quote HistoryQuoted: The difficulty in figuring out what is real and what is fake will be effectively impossible if you didn't witness it yourself. Confirmation bias ftw The media has been selling lies ever since there was a media outlet. It also helps that 90% people don’t think at all. They just go by whatever sound bites they hear in the news. It’s all fake. The entire news media is fake and has been. |
|
|
Win a FREE Membership!
Sign up for the ARFCOM weekly newsletter and be entered to win a free ARFCOM membership. One new winner* is announced every week!
You will receive an email every Friday morning featuring the latest chatter from the hottest topics, breaking news surrounding legislation, as well as exclusive deals only available to ARFCOM email subscribers.
AR15.COM is the world's largest firearm community and is a gathering place for firearm enthusiasts of all types.
From hunters and military members, to competition shooters and general firearm enthusiasts, we welcome anyone who values and respects the way of the firearm.
Subscribe to our monthly Newsletter to receive firearm news, product discounts from your favorite Industry Partners, and more.
Copyright © 1996-2024 AR15.COM LLC. All Rights Reserved.
Any use of this content without express written consent is prohibited.
AR15.Com reserves the right to overwrite or replace any affiliate, commercial, or monetizable links, posted by users, with our own.
