As someone who works with AI a lot, this is the most unexciting explanation of AI I've ever heard. Ali's question was great and I feel like this history lesson is not what he expected to hear :) Reminded me of some of my university professors...
as a AI student, I agree with you. He talked about decision trees and some other stuff like if they're common knowledge... his explainations don't really clear up things and his excourse about the history might confuse people
yep, it felt like a lot of name dropping proving he knows about all these people that we obviously don't know cause we don't know AI (and after this vid we still don't) :|
6:04 it looks like ali is in a brighter environment than the guest. It absolutely disturbs your vision having two different exposures in a same video clip.
For a so called AI expert, I think he explained this poorly. I think he's confused more people than he's helped understand. He's talking about decision trees, as if they're common knowledge. He doesn't define AI in a clear fashion, which is fundamentally what he needs to do because people don't understand what it is. They think they do, but if that were true, this video wouldn't have been released. He starts saying simple addition is AI. In the sense of an algorithm, arguably, yes, but that has no complexity. It's just binary, it doesn't have any iota of intelligence. AI is often related to machines learning and understanding from data/inputs. That's the distinction; with traditional computing, the programmer provides a set of instructions to apply to the data and the computer follows them. With AI, the computer uses an algorithm to identify patterns and generates a set of instructions to follow in order to fill in gaps on future inputs. Humans do the same in a much less precise manner. That's the broad strokes of it. You could argue that that's a little arbitrary, the method to generate that set of instructions still comes from a human currently. So, the human is still kinda generating the set of instructions but just a step removed. Eventually, people believe that humans will be more steps removed, that computers will not need human input. The vast majority of "AI" these days isn't particularly intelligent but it still does extremely cool things. Obviously my explanation still misses out a lot but without a massive comment and examples, that's the most concise version.
I barely know anything about AI but I found this video pretty informative, probably because understanding the history of something helps me make sense of it in general, your comment was also useful but I wouldn't have understood it without the video, so I guess it wasn't so bad. I'm also relieved to know that humans are still doing a big chunk of the work 😬
@@roktaqi It tells you very little about what AI does and how it works. No offense, but you don't know what you don't know. When most people think of AI they think of ChatGPT, image generation models and the like. They tend to be more advanced examples of AI. So when people think of AI it can be scary, but the majority of AI you experience is relatively mundane and you probably don't even realise it. It classifies things, groups things together, makes forecasts, and so on. You might think, "well, how does that help?" But a lot of tasks can be broken down into a variety of these types. Like I said, it's hard to explain without examples. Ultimately, it's statistics. That should have been his starting point. It's also not everywhere. So many people are convinced it is because they don't understand it, but also because they're not familiar with the field (not that I'd expect them to be). I don't think a history lesson is particularly useful tbh. How does that impact your understanding of what it does? It doesn't unless you want to get into the philosophy (and ethics) of AI, which is all well and good WHEN you understand what it does.
Garbage in, garbage out. That's the analogy not just for AI, but for any "expert" like this guy on the interview explaining how AI works. There's still human curation for "proper data" taken into account for the rule generator aided by transformer models, but considering how the internet is filled with garbage data and garbage human brains curating the data, our jobs are safe.
100%. This "expert" is useless. History and development is largely irrelevant to the layperson. It would have been far more prudent to give an overview of different types of AI at the very least
Summary by ChatGPT: In this discussion, the speaker provides an overview of the development of artificial intelligence (AI) and its historical context. The conversation covers early ideas of anthropomorphized beings and the quest for machines that resemble humans. The narrative traces the evolution of AI from Alan Turing's concept of a universal computing machine in the 1930s to the introduction of the Turing test in the 1950s. The speaker highlights the shift from explicit rule-based programming (Good Old-Fashioned AI) to statistical machine learning methods, emphasizing the effectiveness of the latter. The emergence of deep learning, led by figures like Jeffrey Hinton, Yoshua Bengio, and Yann LeCun, is discussed, marking a significant turning point in AI around 2012. The discussion touches on the reframing of AI, moving from explicit rules to implicit learning from data, ultimately leading to the AI revolution.
As someone who works with AI a lot, this is the most unexciting explanation of AI I've ever heard. Ali's question was great and I feel like this history lesson is not what he expected to hear :) Reminded me of some of my university professors...
Yeah.. its not even bearable in 2x speed
as a AI student, I agree with you. He talked about decision trees and some other stuff like if they're common knowledge... his explainations don't really clear up things and his excourse about the history might confuse people
yep, it felt like a lot of name dropping proving he knows about all these people that we obviously don't know cause we don't know AI (and after this vid we still don't) :|
Why does the camera stay focused on Ali? This is very annoying when the interviewee is speaking.
I am pretty sure it was a mistake done by some new editor who doesn't understand the context well enough
The grading between both is very different- and background noises are not removed - too quick edit
It is to show his response and reactions :)
@@mistermotokithe sound is horrible 😂
Love your content always, but what is wrong with your camera, I checked my laptop's brightness twice :p
6:04 it looks like ali is in a brighter environment than the guest. It absolutely disturbs your vision having two different exposures in a same video clip.
word Robot comes from Czech language, mentioned author is Karel Čapek - book name: R.U.R. 🤩
Karel Čapek was his name, not Karel Kapech.
Nice to watch a video that is explaining it in a comprehensive way. I like it.
For a so called AI expert, I think he explained this poorly. I think he's confused more people than he's helped understand. He's talking about decision trees, as if they're common knowledge. He doesn't define AI in a clear fashion, which is fundamentally what he needs to do because people don't understand what it is. They think they do, but if that were true, this video wouldn't have been released. He starts saying simple addition is AI. In the sense of an algorithm, arguably, yes, but that has no complexity. It's just binary, it doesn't have any iota of intelligence. AI is often related to machines learning and understanding from data/inputs. That's the distinction; with traditional computing, the programmer provides a set of instructions to apply to the data and the computer follows them. With AI, the computer uses an algorithm to identify patterns and generates a set of instructions to follow in order to fill in gaps on future inputs. Humans do the same in a much less precise manner. That's the broad strokes of it. You could argue that that's a little arbitrary, the method to generate that set of instructions still comes from a human currently. So, the human is still kinda generating the set of instructions but just a step removed. Eventually, people believe that humans will be more steps removed, that computers will not need human input. The vast majority of "AI" these days isn't particularly intelligent but it still does extremely cool things.
Obviously my explanation still misses out a lot but without a massive comment and examples, that's the most concise version.
I barely know anything about AI but I found this video pretty informative, probably because understanding the history of something helps me make sense of it in general, your comment was also useful but I wouldn't have understood it without the video, so I guess it wasn't so bad. I'm also relieved to know that humans are still doing a big chunk of the work 😬
@@roktaqi It tells you very little about what AI does and how it works. No offense, but you don't know what you don't know. When most people think of AI they think of ChatGPT, image generation models and the like. They tend to be more advanced examples of AI. So when people think of AI it can be scary, but the majority of AI you experience is relatively mundane and you probably don't even realise it. It classifies things, groups things together, makes forecasts, and so on. You might think, "well, how does that help?" But a lot of tasks can be broken down into a variety of these types. Like I said, it's hard to explain without examples. Ultimately, it's statistics. That should have been his starting point. It's also not everywhere. So many people are convinced it is because they don't understand it, but also because they're not familiar with the field (not that I'd expect them to be).
I don't think a history lesson is particularly useful tbh. How does that impact your understanding of what it does? It doesn't unless you want to get into the philosophy (and ethics) of AI, which is all well and good WHEN you understand what it does.
Garbage in, garbage out.
That's the analogy not just for AI, but for any "expert" like this guy on the interview explaining how AI works.
There's still human curation for "proper data" taken into account for the rule generator aided by transformer models, but considering how the internet is filled with garbage data and garbage human brains curating the data, our jobs are safe.
100%. This "expert" is useless. History and development is largely irrelevant to the layperson. It would have been far more prudent to give an overview of different types of AI at the very least
Summary by ChatGPT:
In this discussion, the speaker provides an overview of the development of artificial intelligence (AI) and its historical context. The conversation covers early ideas of anthropomorphized beings and the quest for machines that resemble humans. The narrative traces the evolution of AI from Alan Turing's concept of a universal computing machine in the 1930s to the introduction of the Turing test in the 1950s. The speaker highlights the shift from explicit rule-based programming (Good Old-Fashioned AI) to statistical machine learning methods, emphasizing the effectiveness of the latter. The emergence of deep learning, led by figures like Jeffrey Hinton, Yoshua Bengio, and Yann LeCun, is discussed, marking a significant turning point in AI around 2012. The discussion touches on the reframing of AI, moving from explicit rules to implicit learning from data, ultimately leading to the AI revolution.
loved it. going to watch the full video🏃♂💨
Karel Capek
He dont even care what he talks about . Yeah yeah speak speak cant wait to upload this video
This is the first video of Ali where I have showed dislike. Good history lecture on AI with no utility of AI.
First comment from india❤❤❤❤❤🎉🎉🎉🎉🎉
Hjbhjjj
Ali did you have a Christmas, looks like he is working 😅
5th comment from Pakistan
2 nd comment from india