I don’t often start threads but this 60 minute section showed me how scarily fast AI is coming on. How it can teach itself to improve at amazing speeds. How it can even make things up and lie. Quoting 5 books it gained information from. (Disinformation) On checking no such books existed. How it teaches itself soccer moves or chess. A friends son worked for “Deep Minds” in the UK. Shades of ‘Skynet’ This video is well worth your time.
When I was visiting my mom last week, she told me about this. Nothing that was surprising (to me at least, and probably several here to are following tech/AI development), and it also seemed like it was an attempt by Google to say "see, our AI works, too." Since their AI is behind, particularly Microsoft's.
I didn’t see that, Yeah, it was google centric but the message over all was important. How fast and how far ahead is AI. They interviewed Deep Minds also and they’ve been at it a couple of decades.
I've watched plenty of movies about how AI would eventually try to kill us all. I never thought that the future I should have been concerned about (at least in the immediate future) was from the Al Pacino flick S1m0ne. I figured that the CIA and other government agencies would have the technology to create realistic pictures or video of anyone doing anything. I never considered it would be available at the consumer level. At some point in the near future anyone will have access to AI that, as long as they have a picture and a voice recording, will be able to create a video of someone doing or saying anything, and it'll be virtually indistinguishable from reality. Some random person at a keyboard will create a video of a political candidate they hate doing something that's so egregious that it would kill the campaign. It can't be something like shooting up a school. It would have to be something that was damaging enough to kill a career, but small enough that it's not inconceivable that the candidate or their handlers couldn't keep it quiet. Considering how the news is now, you'd have some networks saying it's real, while others say it's not. It could get to the point where people don't believe video or audio that's actually real. If you think "fake news" is a problem now... Also, I guess I'll get to look forward to new albums from Elvis, Prince, and any other estate who'd like a cash grab. Maybe Sean Connery can come back as James Bond. That'll get people back in the theaters.
Have you been paying attention to deep fakes? Have you been paying attention to facial recognition? Have you been paying attention to voice prompts on customer service lines, or the chat bots on help sections of websites? I ask because all of those things are AI. The bots Russia sicked on twitter to mess with the 2016 election were AI. It has been clear for years that AI would advance by leaps and bounds in the private sector because it made sense on cost basis. ChatGPT and Midjourney just made things free (well, Midjourney is no longer free). I'm actually excited to see what Adobe does, though they won't be free. In the future? It is already happening. There is a huge issue in India where female movie stars are deep faked on porn actresses and people think it is real. A few years ago, some researcher did a joke with Obama and Bush to show both the possibilities and the limits, and it was so convincing that a clip of one (or both, I forget) went viral because people thought it was real. This is already being discussed. There are already joke examples out there. This is already happening as well. This is, essentially, was CGI is. I know there are several examples, but off the top of my head, this is what happened in Hunger Games: Mocking Jay Pt. 1 and 2 with Phillip Seymour Hoffman, who died prior to filming. Also think of the fight scene in the stairways in Atomic Blonde. It was stitched together via AI (CGI) . 1917 was a series of shots also stitched together via the same method. The technology that everybody is freaking out about - not wrongly - has been around for a while. The only difference is that it was released into the wild and is largely free. Well, ChatGPT did take the next step from what was previous, but it is not new.
Wait until someone in the near future says “I don’t like the way this is going. Let’s scrap this experiment and move on” While a unit with a sensor 100,000 times faster than a human brain receives it. Too many movies, too soon?
I agree with just about all you’ve said there. It the way that AI is learning from itself and the speed it’s doing that, that is potentially scary. Think of another decade, or less where it will be. AI in movies and gaming has been pretty juvenile so far. Jet fighters with missiles 100yds apart. Schools of submarines underwater doing the same thing. The movie 1917 looked far removed from the war zone of that time. It works well for viewers who are excited by that but don’t really have a sense of what it’s really like. I’ve watched the first half hour of Saving Private Ryan 3 times but only the full movie once. The rest was just Speilburg assisted AI battle pap that looked like gamer crap. AI in real life though is moving faster with more applications in a lot of fields other than movies. One example shown in the 60 minute was in medicine. Then as you pointed out the bots and imaging. in truth only the immediate applications will bother me, I’m an old bugger. But I have grandkids all in the early 20s, one 21 last week. They’re the ones I worry about. That statement is really quite profound. “As good or evil as human nature allows!” We’ve already witnessed what human nature can do. Good and Evil.
The idea of evil AI killing or enslaving humanity reminds me of the debate over how there was a risk the Swiss Hadron Collider could create a black hole. It's not possible to train an AI to become fully sentient. That AI can train itself and make itself appear like it can think on its own is as far it goes. The only binaries AI will ever know are 0s and 1s, not good and evil. It will never be HAL, Skynet, or the Matrix. It's fun science fiction idea, but not something to be scared of. AI is a tool. To what end people choose to use can be concerning. If there are people who choose to use AI for wicked reasons, then we're talking about a serious matter. But IMO there is far greater potential for AI to be used for our benefit, i.e., for productive reasons, then to our detriment, i.e., destructive purposes.
I get that, but that's still superimposing someone's face on an existing video, or a researcher using his expertise to create a fake. Anyone now can go to an art AI and type in some keywords to get a picture. At some point soon some guy with no computer knowledge will be able to type in "Joe Biden; 15 year-old; sex" in a form and have an AI spit out a realistic, unique video in seconds. That's what scares me.
Probably old news for some but the implications for future dogfights are there. Wingman now, leader in the future. https://www.popsci.com/story/technology/boeing-loyal-wingman-drone-first-flight/
Now the NYT gets nervous. https://www.nytimes.com/2023/05/01/...ytcore-ios-share&referringSource=articleShare
He’s mostly talking about ChatGPT. 'Godfather of AI' quits Google to warn of the tech's dangers Geoffrey Hinton, who created a foundation technology for AI systems, told The New York Timesthat advancements made in the field posed "profound risks to society and humanity". "Look at how it was five years ago and how it is now," he was quoted as saying in the piece, which was published on Monday. "It is hard to see how you can prevent the bad actors from using it for bad things," he told the Times. In 2022, Google and OpenAI -- the start-up behind the popular AI chatbot ChatGPT -- started building systems using much larger amounts of data than before. Then a lot more in the link.
Who wore it better? Waterford Whisper News BREAKING: ‘Godfather Of AI’ Assassinated By His Smart Fridge
More work for your chatGPT. Can chatbots handle medical questions better than doctors? Study says yes SAN DIEGO -- What are my odds of dying after swallowing a toothpick? Do I need to see a doctor after hitting my head on a metal bar while running? Am I likely to go blind after getting bleach splashed in my eye? A new study led by researchers at UC San Diego explores how artificial intelligence compares to human expertise in the workaday task of dashing off quick responses to routine medical questions. (And more) https://nordot.app/1026407701531115520
We should stop calling it artificial intelligence. It's not intelligent. It's simply fast in doing what's already there, the fastest copycat.
Well, AI is doing good to the stockprice of ASML since NVIDIA told they can hardly manage the demand for chips because of AI.
Looks like a lot of AI folk are getting nervous of Skylab. A friends son works at Deep Mind in England is getting nervous. I took one section from a long report in the link. A.I. Poses ‘Risk of Extinction,’ Industry Leaders Warn Leaders from OpenAI, Google DeepMind, Anthropic and other A.I. labs warn that future systems could be as deadly as pandemics and nuclear weapons. Some skeptics argue that A.I. technology is still too immature to pose an existential threat. When it comes to today’s A.I. systems, they worry more about short-term problems, such as biased and incorrect responses, than longer-term dangers. But others have argued that A.I. is improving so rapidly that it has already surpassed human-level performance in some areas, and that it will soon surpass it in others. They say the technology has shown signs of advanced abilities and understanding, giving rise to fears that “artificial general intelligence,” or A.G.I., a type of artificial intelligence that can match or exceed human-level performance at a wide variety of tasks, may not be far off. https://www.nytimes.com/2023/05/30/technology/ai-threat-warning.html
The fact that after all of this time A.I. hasn't taken over porn makes be skeptical that it could ever be dangerous enough to be an existential threat.
Just human like. If you don’t know then just make it up. https://www.theguardian.com/comment...rch-avianca-statement-ai-risk-openai-deepmind