Will AI Make You Obsolete?
HTML-код
- Опубликовано: 2 мар 2023
- Start speaking a new language in 3 weeks with Babbel 🎉. Get up to 60% OFF your subscription ➡Here: go.babbel.com/t?bsc=1200m60-y...
Let’s talk about Artificial Intelligence, what it is, how it works and where it might be taking us. We’ll try and see if as an investor you should worry about missing the boat on AI, and as an individual should you worry about losing your job to these new technologies?
Patrick's Books:
Statistics For The Trading Floor: amzn.to/3eerLA0
Derivatives For The Trading Floor: amzn.to/3cjsyPF
Corporate Finance: amzn.to/3fn3rvC
Patreon Page: / patrickboyleonfinance
Visit our website: www.onfinance.org
Follow Patrick on Twitter Here: / patrickeboyle
Patrick Boyle On Finance Podcast:
Spotify: open.spotify.com/show/7uhrWlD...
Apple: podcasts.apple.com/us/podcast...
Google Podcasts: tinyurl.com/62862nve
Join this channel to support making this content:
/ @pboyle
Start speaking a new language in 3 weeks with Babbel 🎉. Get up to 60% OFF your subscription ➡Here: go.babbel.com/t?bsc=1200m60-youtube-patrickboyleonfinance-mar-2023&btp=default&RUclips&Influencer..patrickboyleonfinance..USA..RUclips
Not gonna lie, I was expecting the example phrase to be "Una cerveza por favor" when you said you were learning enough to get by in Spain.
On a side note: Great work on the lightning !
Una mesa para quatro por favor.
(check out Kevin Bridges for the relevance...)
Vinograd - vee-no-grad.
I want a Boom boom Chow AI.
AI in business is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it so everyone claims they are doing it.
And the people doing it the most are about to ruin their lives
@@ericrowe2533 lives may be a stretch. Credibility and credit rating definitely
.....and those who ARE doing it are not doing it properly!
Great metaphor for so much of what comes out of biz these days.
@@annjuurinen6553 I would love to say it's mine but I've repuposed what Prof. Dan Ariely said about 'Big Data' a decade ago.
Patrick is beyond under rated. Setting the stage for new RUclips. Thank you for no flash, no bs, and helping us. Grace to you mate. Cheers.
good books as well. have his statistics for the trading floor and pricing financial derivatives books.
Imagine discovering him at the same production and content quality back when he was below 100k subscribers.
Maybe we should see if the AI invest in crypto? To determine intelligence.
He's really good. But if RUclips just becomes people reading out their essays then it will get old really fast.
That's what the wireless was like a hundred years ago.
Partially true.
It's usually great as long as he's not going into details, particularly in case of technical stuff.
Even this video was on the edge and he was carefully trying to avoid any technicalities, that would expose him for not understanding technology.
Also, he is blinded with hatred towards Elon and is willing to twist and fabricate things just to be able to mock Elon, Elon's projects and anything, that Elon speaks favourably about. Even if this cause video to end up being detached from reality both in numbers, situation description and technology consequences…
I'm here for the sarcastic roasting
🤔
@@bossshoang there wasn't much sarcasm today, but that's okay
Same
I am 18 mins in and not one snide Kevin O’Leary comment so far. Disappointing!
Did catty woods hurt you 😂😂😂
Since ChatGPT has started writing Patrick's scripts, the videos have become longer than they used to.
There was a time when a job "nobody wanted", such as cleaning lady or mail room clerk, could serve a stepping stone to moving up the corporate ladder. The disappearance of such opportunities is the real tragedy of automation. One example would be Gail Evans who started as a cleaning lady at Kodak, and another would be David Geffen who started his career at the William Morris Agency mail room.
Upward mobility from these positions began to be curtailed before automation was a factor. It was, rather, the increasing dominance of education ideology, or the notion that every position above that of a shitkicker required a degree, that ended upward mobility within companies. It's not that these jobs disappeared but that companies stopped promoting from them. On the job learning ended because everything, including training, was outsourced.
Or Saul Goodman, who started in the mailroom at the H.H.M law firm
Well I have lived through such a shock as you mention. When the Iron curtain fell, over 80% of all employees in East Germany lost their job within 2 years. It was absolutely insane. And yes, most found a new one. But only most, not all. The outcome was still devastating. Almost a third of the population left the area. The younger third.
Not even a comparison, you dimwit. Rescind the right of thoughts, brainlet.
AI is bullshit. It's just a very advanced grammar checker. We need to stop placing any value in it its just another dumb computer program. It understands nothing and creates nothing. Dont worry about it is pathetic, let the young dumb down further, we need more stupid slaves to look after us in our older age, so let this Ai create a dumbed down race a ignorant slaves, we will be like geniuses to them.
But solution is not to keep iron curtain.
@@ivani3237 oh, of course not. the country as a whole is sure better off now, but the cost was high.
@@ES-sb3ei Ha, East Germany had the smooooothiest and easiest transition from any other post-socialist country.
This is by far the best rap-focused channel on RUclips!
Thanks Patrick. Wise words. My coworkers and I heavily automated our own jobs (engineering) decades ago. Nobody got fired, we just worked 4-5 times faster, created time for innovation, customer support, operational support, sales support, some integration, etc. All good things.
AI is just a tool, after all. An optimistic look is getting rid of all the boring jobs for better things.
I think it's crazy that we are asking if computers are intelligent when we can't even roughly define intelligence.
But about working hours?
@@iraklimgeladze5223 he works a lot more for the same pay as someone who didn't automate anything.
@@zesky6654 yes, he mentions that he increased productivity but not mentions if that increased his pay or decreased working hours. What benefits he got doing that much?
In US tech industry investment, it's better to be wrong than be late. So every time a new buzz word comes in like cloud, VR, ride-share, block chain, AI etc, the hype money starts to stir and devil take the hindmost.
I find that a bit worrieng, since there doesn't seem to be any space for the question 'should we actually do this?' At the moment things still might not seem that alarming, but with the speed at which things change, this question seems more and more important to ask
@@henrikejekel2247 Humankind exists to make progress.
That is a good point.
Problem is. So many investors fall for scams by just jumping into any company that has CGI, and a solid pitch.
AI is a little different. It will happen, but the time line is all over the place.
Improvement of robotics that is. Not full fledged human intelligent AI.
The thing is. Even if it happened tomorrow. It is going to take generations I think before they completely take over everything.
Look at how many advances actually didn't get rid of the things they are supposed to. From Fax Machines to Tax Accountants with computers to Nuclear for power.
I think the last few generations will try to push for legislation to keep jobs for the sake of saving jobs. Until we realize it is putting America behind to countries who embrace it.
Warren Buffet says if we taxed corporations 21%? That would pay for our current budget. So taxing at some rate like that while giving UBI will be the future. Who knows how far off. After a civil war most likely.
Don't think I've come across a channel that provides such quality content. You have a deep understanding and provide your insights in a concise way
as somebody with PhD in CS (focusing on ML) and currently working in R&D for the industry, I must say this is an amazing overview - thanks Patrick!
Buzzfeed: We're gonna use AI to make content.
Proceeds to fire the entire writing team.
Content: improves
@@hankhillsnrrwurethrafacts. Really, will be better.
Not the whole team, but it can produce quick answers which can then be edited.
Excellent content! As a software developer I can say that even though you're not a tech guy, this video puts realism into a topic that mainstream media loves to fantasize. I really enjoy your videos! Thank you!
Patrick you are by far one of the most intelligent and well sourced producers of RUclips content. I always walk away more knowledgeable on the topic being discussed. Your depth and research is impressive and your sarcasm greatly appreciated. Thank you!!
Economics, History & Philosophy with a healthy dash of sarcasm. What a great presentation.
I came here for entertainment - to listen to Patrick’s epic sarcasm and how we can still keep a straight face 😂
"AI" is (again) what "blockchain" was to wall st a few years ago
You sir are amazing, and amusing in an intelligent way. Thanks so much for the enlightenment again.
Creating more jobs we actually enjoy 😂 Burger flipping, sewage cleaning, new haircut, getting Botox. Lots of things we are happy to do instead of filling out forms.😅
i've never been worried about ai actually supplanting humans at anything, but literacy and attention span are nosediving these days and i'm pretty worried about the effect chatgpt will have on people who aren't up to the task of noticing it's not a person.
I'm sorry, your comment was too long, can you summarize it in 10 words or less?
@@Falconlibrary comment too long. 6 < words pls.
Exactly. Imo we are becoming increasingly impatient. With short msg, 5 second ads, short news segments(we're here with dr so-and-so, a specialist in x... I'm afraid that's all the time we have), and the even the cutting of films. You rarely see a person walking out of the room. You see them start walking and the cut to door closing. It's remarkable when you reflect on it.
Attention span hasn't been nosediving. Maybe yours has.
@@nunyabusiness863
No, we're not. The ads are getting longer and you can't skip them.
Just a small elaboration on the Getty Images logo thing since there are some common misunderstandings about it:
Stable diffusion models have been trained on hundreds of millions of images and learn to "expect" things like watermarks, logos and signatures. When the model is used to create images, it has no access to any particular image, but instead creates things out of random noise that looks like the things it has come to expect. So it make scribbles that looks like signatures and it makes things like logos, but these are not copied from other images and mashed together as so many have understood it.
You can run Stable diffusion on a computer without internet access, and the file containing the model (or more accurately: model weights) is just a handful of GB large at most, so there's no space for those hundreds of millions of actual images to copy from.
Fantastic video, Patrick! Thanks for putting everything into perspective!
I love this channel. Patrick, you're a true gem.
i like how you used AI to create the poster image for this video, but the AI kinda screwed up the monitor part so you fixed/adjusted it today, very meta comment on the situation with AIs
One of the biggest underappreciated hurdles I suspect AI will face is in actually integrating into regular businesses. Using ChatGPT on its own is one thing, but to actually replace most knowledge work would require the AI to have access to, be able to comprehend and helpfully interpret novel company data, usually spread over a variety of mediums (e.g. emails, databases, pdfs...etc.) and not always well organised, not to mention understand subtext from meetings...etc. There is a lot of architectural engineering that would have to go into that in most companies looking to seriously adopt AI, not to mention companies will have to be _certain_ it's secure.
Until that actually happens, I suspect actual productivity gains in most jobs from AI will be marginal - documents will be quicker to read and write, spreadsheets will be autofilled, coding bugs will be identified...etc.
The way I see it, we can't even use the limited AI we have to it's full potential. Meaning that it's already us who is slow to adapt, and not the AI.
@@julius43461
No, it's the AI too. It's overrated.
@@bobfg3130 After using it for work a lot lately, I have to agree. I tried creating copy for a product listing with it, and although what it spits out looks amazing on the surface, there are clear signs that it has no idea what it's doing. Took me so much editing to get it to work.
I do find it immensely useful for writing text where keywords and character limits are not an issue.
The best part of this video was when Patrick said the word "Investors" & an image of three guys in North Face vests drinking Starbucks flashed on the screen.
You have earned a Patreon subscription. The info, the precision (I teach philosophy of AI), the humour, and the Dublin accent that makes me homesick for my 6 years in Dublin. Cheers!
Interesting fact - The first use of artificial voice generation was used in the movie 2001 Space Odyssey which interestingly came from Bell Labs and one of its inventor John Kelly of the Kelly Criterion Betting scheme.
Patrick, your video is well thought out. As an AI practitioner of many dog year who has lived through "AI" winters, i agree with your analysis. Now if you could get Google to use "AI" to improve its its spammy ads on your video, we would all be better off😂
Fantastically explained! I work in this field and Patrick is spot on with most of what he said.
I love you channel and content. I just wanna say your really funny content about the "pumping and dumping" needs more. Keep doing what you're doing sir, it's much appreciated!
Thanks Patrick, another great chat
What a spectacular video. I really enjoyed that…and feel more informed as well.
(Also…the blistering sarcasm had me lol’ing.)
Love your humor filled narrative work.
Thank you for another great video mate. I enjoyed every second. Please keep it coming.
Love your work Patrick, thank you!
This is my direct field as a data scientist and I think you just explained it all better than I do. I’m now concerned you are coming for my job!
If there's one thing AI won't do it's sit in court after an accident happens. From bus drivers to pathologists that's a big problem nobody seems to mention when it comes to actually replacing the employees that enthusiasts say this technology will soon replace.
This is probably the biggest short term risk with AI as I see it. Not that a super-intelligent Skynet will conquer us, but that dumb humans will put dysfunctional AIs in positions of responsibility it has no business being in.
Thank you Patrick. This is such a insightful video
Best content I have seen on this subject, since I First heard about GPT about 8 months ago.
The thumbs up was well deserved👍
This video is fantastic quality, well structured, impossible to replace with AI.
0:20 The moment you mentioned "perceptron", I knew your video was going to be a stinging indictment 😄
Seeing Professor Boyle talk about my field is awesome! I do hope that some good uses will remain once the hype wears off.
My favorite types of media coverage of chatGPT is when scheisters use leading questions and strongly influence results then act that the tool is more powerful, sentient, or intuitive than it really is.
Skyientists are making outlandish claims.
Hahaha only two minutes in and I’m already laughing 😆 Causal Cathie Wood reference got me 😂
Pressed Like ahead of time! 🔥
Thank you for great vid presentation. Truly enjoyed as the actual technology scares me greatly. See you again soon.
Although not a finance person I always learn something watching Patrick’s videos. They always seem to fill the gaps the rest of the coverages leave out…
Excellent video, as always
The hype cycles are getting so fast, each year there is at least 2 hot topics that the whole world jumps behind.
Thanks Patrick for the great video.
You are a delight. Your pithy commentary fills me with joy. Your assessment is so very clear eyed when it comes to AI. Thank you.
EXCELLENT PRESENTATION, AS ALWAYS, PAT.
If I took a shot every time I heard a techbro say that technology always creates more jobs than it destroys recently, I would be dead.
More dead than most people's careers in a decade.
I would fear even more than skills replacement, about being overwhelmed by offers tuned to current top clickbait - with AI building content from public mess and biases. How you will find a high quality offers?
Absolutely brilliant video. Nice philosophical touch there
Love love your videos!
Patrick Ive been subscribe ever since the first video with Coffeezilla. Id just like to say thank you! As someone who is just an average 9-5 kinda guy you have broken down soooo many concepts in finance and business for me. This video in particular has quelled many worries that i had about this "NEW" AI technology and i have a much better view and understanding of it. So once again Thank you and more power to you man
Kurt Vonnegut wrote about an automated world in his book "Player Piano". It's a fun read.
So it goes.
Love the analogies, and couldn't have stated any of this better myself.
A really great video bringing people down to earth about this.
Just excellent explanation
I like the calculator analogy. That really helps put automation/AI into perspective.
Peple without calculation skills can function in society thanks to machnie calculators. People without language skills can now function in society thanks to language models. Someone interested in journalism but lacking the writing skills can now become a journalist anyway.
this is the best peice on AI i have seen anywhere finally someone who talks about skills instead of jobs and puts the development into it's proper perspective.
I have never had someone calm me down about something with facts and logic that fast, thank you for this video
That's one very thorough and objective summarization of the current situation about "AI".
Thank you for providing us such content - much appreciated!
The twelve seconds we'll have left when people are forced to learn the difference between "AI" and AGI are gonna be weird
@@boldCactuslad Expert say it will be a continuous process, not just a single moment when AGI starts to act consciously :)
_Man, that introduction was absolute fire._ 🔥
Who knows, maybe even the *Neanderthals* had a revolutionary *AI* in the works, unfortunately they brought their secrets with them to their grave.
Every Friday - Boyle! Well done as usual !
Bravo, Professor. One of your best!
It has been sooo frustrating to watch so many people rave about AI, and more annoyingly, often get hostile when presented evidence that "these are cool breakthroughs, and powerful tools, but the hype is far beyond that"
I frankly put off this video at first because I was really worried this would be yet another "Here's all the sectors that will be disrupted by AI!" thing, so thank you for giving a detailed overview.
I've always found one of the best examples of the limits of AI's to be video games. An amateur can build an AI that will beat every player on the planet in fighting games, as it gets to leverage inhuman reaction times to play perfectly.
And yet, there's no AI that can beat something as simple as Mario 64, a game a child could beat. If you want to drop out of 3d, I've not even seen an AI that can handle something like Mario 3 or 2, and while I have seen "AI's" that can beat mario 1, its by brute force. Literally simulating thousands of games at once with random inputs and pruning outcomes/reloading previous states until it gets the outcome it wants.
And that only works with mario 1 because its a "go right" style game. The later games aren't being beaten end to end by AI because AI has a major problem with exploring. It doesn't know when to go left or up instead of right.
The tech is still fascinating and even world changing, but we're far from fully self driving cars for example. The amount we could automate RIGHT NOW, if we just got everyone on the same page, is huge. Imagine how much easier it is to automate a train than it is a car on a road? This same pattern is all over the workforce, and doable with super simple automation, yet we still constantly buck away from standardization and wind up with more and more human input required.
😌 I am a Machine Learning Engineer and here for the roasting of the public (media). Thank you.
it should be a roast of social media and the parasocial relationship exploitation. people dont care about AI actually at work for them that we already have. People are obsessed about something that feels like theyre talking to something, using the parasocial relationship so many fall for. chatgpt uselessly "talks" and so people love it. Customers mean money. Thats it. Its idiocracy.
Roko's Basilisk has taught me to be thankful for your efforts. Nothing but fear and respect coming from me my dude. Thank you and don't hurt me.
Fund managers and investment advisors are an easy 1st hit for ai. Me thinks Paddy's youtube career is a smart move if the hip hop or rap gigs don't work out.
thanks for the reasonable volume of the outro tune!
Excellent video Patrick! As a computer scientist I subscribe everything you said, specially that jobs that require human intervention will stay that way. Face to face jobs are necessary.
You mean like people taking orders at fast food restaurants? Or bank tellers? Or salespeople? Or travel agents?
@@rdean150 Yes, all of them. Think about old people that do not know how to operate a simple UI, they need somebody to help them. Banking is having a serious issue with that in my country (Spain), they are reducing extremely the staffing to cut costs and do it all online, only for the online services to not work when you need them. You end up having to talk to a real human. Customer service ain't going away any time soon.
@@MrMCMaxLP Ah well man of those jobs have already been eliminated and replaced with technology in the US.
Even if it causes problems, companies are still doing it.
@@MrMCMaxLP He basically listed jobs that no longer exist.
Excellent video as usual. The point about the ancient Egyptian writing for surgeons reminded me of the recruiting process that the Lyon's Tea Company implemented when hiring engineers to program the Lyon's Electronic Office (LEO) - the world's first dedicated business computer back in the 1950s. Since programming was an entirely new field, they tested applicants' logic skills by asking them to create a flow chart for making the perfect cup of tea!
thank you, great assessment!
What a treat, thanks!
Begins and ends with a jab at Cathie Wood 😂 A+ content from Patrick, as always
😂
I would not listen to her for nothing. She looks like a crazy!!!
Thanks for this video Patrick!
As someone that has been working with primarily machine learning for the past decade, it's good to see a well-researched and presented piece like this.
The vast majority of our clients have dealt with industry-specific improvements, be that in monitoring and alerts, or data analysis to present better foundations for business decisions.
None of our clients have sized down their workforce in the areas we have delivered projects for, if anything they have hired extra people and they can now enjoy several clear %-age advantages over their competitors.
The best results we have seen is to use AI/Machine Learning to give better tools for human operators to use, or better information on which to base decisions.
Thank you, Patrick. Again a great video. I always gain new insights watching them. - but what about the wood nymph?
Articulate , constant. No wasted words. Funny when able and poignant. Your best video yet. And certainly an important subject.
"I'll be back".. Best AI line in a movie... Second to ... "Do you want to play a game ?"
This is the most well-researched videos I've seen outside the computer programming industry.
One of the best thought through commentaries on 'AI' that I have heard. Comments on rate of change and distribution of impact are especially trenchant.
Buena suerte, Patrick! Otro gran video.
God bless you, Pat. I'm heavily into Alphabet and I must confess, when I first heard of GPT, I was worried. Yet, now I realize this may write a bad poem. But, it won't disrupt search, anytime soon. Always love you for your entertainment qualities!
Have you tried Bing lol, it beat google search in every single way. If google is really prepared for AI war they wouldn't have such a terrible demonstration of Bard. The reason Google is so slow in integrating AI in search engine is because the majority of their revenue is from ads, they are already doomed because MS took all the spotlight and progress. Now they have to chase behind MS for 1-2 years of development
Thanks to our growing list of Patreon Sponsors and Channel Members for supporting the channel. www.patreon.com/PatrickBoyleOnFinance : Paul Rohrbaugh, Marc De Mesel, Nate Stapleton,Timothy Baird, WIlam, Hernan Merino, Random Encounter, Nieuwsbrief Ikwil, Bee Positive Consulting, hyunjung Kim, John Cadena, Ian Tracey, Callum McLean, Oscar, Simon Pena, Ed, Pavle Obradovic, Erik Van Ekelenburg, David O'Connor, Pjotr Bekkering, Alex, Robert W Proudfoot, Robert Muller, Andre Michel, Ivan Iliev, Gopaljee Atulya, Mark Hooker, Artem Vasenin, P H, Sebastian, Michal Lacko, Peter Bočan, Michael Pierce, V Jordan, Gil, HalfwitHam, Mark Brophy, David Urdenata, Juan Valdez, Bruce Roberts, Chad Norman, Bruce Roberts, Shamikh Rana, Friday Guy, Marc De Mesel, Augusto Ramos, Soy Boomer Doomer, Bob Slartabartfast, Robert Feiler, Camil Dbouk, Erik Montesinos, Matthew Loos, Az Indragiri, Aman Bali, Lautaro Parada, Pratap, Deborah Joseph, Robin Sung, Kurt Johnston, Dominik Auerbach, Gurmeet Kaushal, John Hall, Dara Mo, Josef Goergen, Wilbert Cheng, Jaroslav Tupý, Trevor Lucey JB Weld, Alex, Carlos Figuera, Peter Pomelov, Null065, Rick Thor, MeBerzerk, Henry Nguyen, Sola F, The Collier, Carlos Mejia, J Wadia, Bitcoin OG, easy boekhouding, Albert, Eugene Jung, Daniel Cervini, Jonathon Yong, Iris Ji, Emil Nicolaie Perhinschi, Charles, Eli Auto, Excks, Michael Li, Par Hedman, Praveen Mishra, Gerard Scott, joel köykkä, Areeb Ahmed, David Wang, Yazan Qaraqish, Rodolfo Cornetti, Daniel Winroth, johnny, Nick Jerrat, Chris Houston, Alastair Currie, Robert Griffin, Andrei, zizi Golo, Fab Vida, Constantin Petrenco, pawel irisik, NotAScam, James Halliday, 22 Dust, Carsten Baukrowitz, Heinrich, Arron T, Ben Brown, Stephen Mortimer (to The Moon), Ryan B. Hicks, Liam, Logan Vrankovic, William Heaton, Paul McCourt, Daniel, Aaryan Koura, Steven, Christopher Boersma, Ulf Lundblad, Dorothy Watson, Greg Blake, Simon Bone, Livermores Quant, The Collier Report, Scott Gardner, The Man Koala, Brian McCullough, Daniel S. Smith, Finance Student, Julie, Mohammad Rehman, James Wallace, Daniel Poellmann, Edosa Odigie, Dixon Yuen, Marek Novák, Stamatis Drepaniotis Michael Smith, Ahmed Hamadto, Chris Davey, Mike Farmwald, Michael A. Mayo, Lachezar Georgiev, Kamet Batra, Bradley Johnson, Sagar Gudi, Michael Chessar, Kate ATL, Tong Cheung, Lady Dje, James Barnes, Chris Hall, Kurt Johnston, ICBM Catcher Juan Valdes, Linn Engström, Veltsh, Konrad P-kala, Pastacat, Adam Vorting, Matthew McQuade, Christopher Lesner, freebird, Kenneth WedMore Lund, erfective, Jason Young, Jonathan Kopnick, Peter Hendrickson, steel, Bastien, Tom Willett, Chris Whitehead, Anil Jason, JOJO, AS7, Greg Thatcher, Ezekiel Templin, MrLuigi1138, Leszek Frankowski, Nam Nguyen, Karim THIBAULT, C, David R. Ingemi, Robert Wave, Dmitri Alexeev, Aaron Rose, Ethan Hernandez, Claude Chevroulet, Stephan Marosvary, Louis Julien, Jan Lukas Kiermeyer, Gearoid O Connor, Fredrick Saupe, Subliminal Transformation, Alex McMahon, Adi, Ben, Kurt Mueller, Janusz Wieczorek, Federico Viscomi, Corgi, Mahdi, Burgerinn, Quinn Cone, QiKaiQian, Stephen, Joshua Rosenthal, Frank Yashar, Michael Smith, Emilian Marius Tudor, Julian Aßmann, Cormac, Ian Shearer, Michael Kopřiva, Tinni, Goran Milivojevic, Joe Del Vicario, Alexandre Mah, ultima9, Norman A. Letterman, maRiano polidoRi, Stephan Prinz, Gary Yrag, Mattia Midali, Matthew Berry, Ann Williams, Jay T, Gabor, Florian Haas, Shivendra Saklani, Zachary Tu, Jeffrey, Lane Alan Deyoe, Chett Flynn Jonathan Horn, Mo Herbert, Lane Andrews, Justin Thuet, Olaf Thiele, Ivan Ilaev, Todd Gross, Douglas Caldwell, Wade Hobbs, Volodymyr Palii, James Hoctor, Gavotti SGP, Ryne Davis, Jean-Philippe Lemoussu, Keanu Thierolf, Muhammad, Michael Chow, Stefan Alexander, Miroslav Ognyanov, Scott Guthery, Vanya Davidenko, Arto Karhu, James Bache, Jason Harner, Dale Patch, Stefan Penner, Mischa Trunz, Arvid, Eric, Jonathan Metter, Junhong Low, John Way, Maria Baker, Luke Solomon, Sharath Vulupala, Keith Elkin, Chris Nicholl, Luis Carmona, Vinci Chan, Olivier, Yasha, James Yoh, Eduardo Martinez, Adi Blue, Swain Gant, georgejr, hyeora, old gambling art bag, Boussaken, Lukas Braszus, Vik, Chris Albertson, Sprite_tm, matt f, Douglas Caldwell, Adgn, Ronald London, Chris Rock, Tuan Nguyen Minh, Daniel Baak, Jeremy King, Julien Debache, M1, Dougald Middleton, Tom, Diarmuid Kelly, Subodh Kafle, Gregory Mahoney, Angelo Rauseo, Kirsten Banuelos, Ryne Davis, Era, Anne Molphy, Ekaterina Lukyanets, Alfred, sugarfrosted, Okkie, Larry, Sarah, miilo, EatEmAll, Alexandre GUYAMIER-CROISSANT, Alex C, Henk S, Noel Kurth, John Tran, Daniel Soderberg, Daniel Ralea, Steve Crotti, Shaun Deanesh, DaFlesh,Nathan McIntosh, Dominique Buri, B P, Manmeet Sheera, Stephen Walker, Richard Stagg, Bo Grünberger, Justin Sublette, Zackk, R, Omega Universas, Paul Twilley, 인기 김, Dennis Arthur Rose, Louis Görtz, Arjun K.S, josef strand, Simon Crosby, Alex Do, Minnah seoh, Ljube Boskovski, Bertrand Barroux, Jacob Snedaker, Edgar De Sola, Izidor Vetrih, WhiskeyTuesday, Vic Tan, Compuart, Felix Goroncy, Nick Top, Anita Mui, C.J. Christie, lazypikachu23, Olivia Ney, Steven, Jordan Millar, Reagan Glazier, Viki Rebic, Adrian, Adam Kobus, KoolJBlack, Deborah R. Moore, Alonso Ibarra, sam, Dru Hill, Dean Maurer, Tim Jamison, gavin, Ultramagic, HEWHOSHALLNOTBENAMED, Nigel Knight, Daniel Winroth, Paul Niekamp, Job Zamora, Zhnjy, Brian K. Lee, Tony Bianucci, storm, scott johnson, joevan burke, Suzy Maclay, Just Plain Old Boring Mike, Fernando Garcia, Agatha DeStories, Chris Peterson, Nesh Hassan, Molly Carr, Christan, Michael Jones, Ross P, Conor Rainey, Milos Krljanovic, Laura Sanborn, Richard Hagen, Nicholas Muller, Adam Stickney, Peter Weiden, Mate Visky, RVM, Brainless, Marcio Andreazzi, Anand Norman, Yih Pin, Ziad Azam, petros ventouris, Justin Zander, Jonathan, STEPHEN INGRAM, Panks, Michael Wilson, Peter Luu, Gregory Lethbridge, Daan Jetten, Ivan Katanic, Benjamin, Boris Azais, Flanneryo, EricO, DoubleWhy, Paul Hilscher, dz, Manuel Barkhau, Gaurav Batta, Earnest Williams, Harun Akyürek, Dionysis Partsinevelos, toufik ouriachi, Maximiliano Rios, Claire Walsh, AT, Atanas Atanasov, William J. Murphy, Tom Eccles, Matthew Lang, Matthew Wehttam, Fernanda Alario, nishil, Alfred Tagher, Ulf Lundblad, Weijie Diao, Dave Jones, Matthew Colter, Alan Medina, Georgios Kontogiannis, M, Kirandeep Kaur, Charles, Jacob Wachala, William Ching, Omer Secer, Clement Schoepfer, Tarek Rabbani, Sofia Tuff, Christopher Brennan, A M, Ivo Stoicov, and Yoshinao Kumagai
I refuse to subscribe to an AI generated Patrick Boyle - that is the hill I am dying on :(
Show’s on!
Your sponsor for this video got a great deal, its content is way beyond intellectual. I wish the main stream media invested this much thought into its content. Very informative and insightful. Mind = blown.
After a long time effort of developing and after some quite big expenses in the field of large language models the main actors came to the conclusion that it is a good time to explore the possibilities of making their product a source of income. That's why LLM is so much present on social media now, much more than before. This can also be a statement that further progress in LLM is not available, what we see is the final product. It reminds me of the evolution of cryptocurrencies before the big crush last year, they also launched a very strong media campaign.
I'll say based on some recent scientific papers ( like the LAAMA model one by META) that AI not progressing further is not true. The research shows that further progress can be achieved by training the model more, or by training it on better data. Yes, it also shows the the progress achieved in such a way is not exponential, is linear and decreases with the more you train the model, but further improving can be achieved, nonetheless. Also this models are not new, the transformer arhitecture on which LLM like chatgpt are based was first developed arownd 2017 if I remember correctly ( "Attention is all you need" is the name of the scientific paper on transformer linguistic models) so is not a new thing, but this things were not in the media ( mainstream, not specific media I mean) until recently. I'll say it's more like they finally launch chatgtp as a product because they thought it's good enough, not that it can't be improved. The last think I wanna say it's that further rapid improvement of this technology can be achieved with improvement to its core arhitecture, but that's impossible to predict when, if and how will happen, but nonetheless slow improvement can be achieved by just improving training, the training data or the size of the LLM in question.
Great to hear Searle quoted on language and meaning - takes me back to grad school sociolinguistics. Searle's point is even more potent when you track it back through the logical implicature: if AI can do some things, but is kind of Calculator for Words and Pics, or just a pixellated difference engine, then what does that suggest for human labor? It means that on the thinking side there will always be roles for human interaction, communication, assessment, judgment, and risk appreciation, no matter the volume or technical difficulty of the inputs - which can now be received, managed and processed so much faster by automation. As a simple example, consider courts administration - incredibly inefficient processes and vast amounts of information, records, and so on - but the people who are fundamental to the task are the ones in positions which require sophisticated meaning-making or people interactions - judges, bailiffs, cops, lawyers, plantiffs, defendants, sergeant at arms etc. The ones who interact, appraise, make judgments and hold accountabilities will be needed long after AI has streamlined every process. In a very real sense AI is going to take away white collar jobs which were invented in the 20thC, to fill an expanding business or organisational niche and are no longer required. I've seen architecture and graphic design already go through a rollercoaster of hopes and disillusionment. But I'm also yet to see the supposed "paperless office", so who knows.
It's easy for us to think that now, because we value that meaning-making work. I would suggest that there is a real potential for a future world in which that work is devalued and is no longer considered worth paying a human to do. Think of those workers as craftspeople - they have unique skills that are currently hard to replicate, and are seen as pretty essential to a number of activities - similar, perhaps, to carvers in stone or wood who were previously needed in the building trades but who have almost entirely disappeared now (when was the last time you saw a gargoyle go into a new construction?). "We'll always need [x profession]" says a lot about our current values, but nothing about future values.
"Communication, assessment, judgment..." You mean like reading through a stack of resumes and identifying candidates who may be a good fit for the role you are hiring to fill? Because that sort of task has infamously been relegated to AI in many companies and the results have been highly controversial.
I shudder to think what will happen when AI is used in our legal system. How many wrongful arrests and convictions will occur because of these probability engines?
@@rdean150 I was hoping I was making the inverse point, but that's an interesting angle. Aside from hiring, how about firing? How does Disney fire 7000 people en bloc? I doubt it's through humans appraising a master list of skills and performance against strategic workforce planning guidance and ticking each one off.
I find Searl's argument utterly unconvcing. My brain consists of neurons and synapses and each of these neurons and synapses have NO IDEA about what is going on. Still, out of complexity of my the interconnections between these neurons and synapses emerges my consciousness.
To believe that machines can never be really "intelligent" is to believe that miracles happen in our brains. Some sort of physics, or magic, that cannot be emulated by machinery (even though we are biological machines).
I do not believe that there is anything "special" happening in our brains apart from complexity.
@@mxvega1097 By delegation.
Excellent. Thankyou 👍
22:15 As a writer, I felt that. As a copywriter, I'm cool. Tailored advertising requires human feedback, period. And that's the last thing humans still have an advantage over AI: tailored experiences.
I work on AI and this video is best take that I've heard from an outsider. Great work, you really put in the work man.
@Premier League you really picked that up from that comment?
Where have I mentioned that I'm special?
Looking for reasons to be offended when there are none. Wow!
get well soon!
Very Interesting, balanced, and full of often uncommon common-sense perspectives on AI, work and investing. Thanks!
Boyle reminds me of my college professors. And for those wondering, that's a compliment. I went to school in the 90s.
Same here, mate! Graduated in ‘96. 🙂
Automating creative tasks is a knife pointed right at the heart of what it is to be human. We're automating ourselves into just being chattel.
I can already see sewer cleaning being a great career in the near future. AI won't be able to automate that just yet.
I'd like to learn C++ or COBOL. Definitely the most accurate video on AI I've seen.
might be worth looking into rust. It's just as fast as C++ but it's memory safe and way easier to write (although the initial learning curve is still steep), it's still fairly new but there's a large community backing it and a number of companies are starting to transition to it from C++
I heard from my IT Professors they are worried about COBOL being not as maintained after newer APIs are being developed outside of banking system and companies will migrate to other flexible programming languages. If you're into deep financial related programming, it's fine to learn COBOL. Else, it might not be a good idea.
Same my understanding is cobol is unlikely to see tons of the same growth because it's rather inflexible
If you're serious about programming pretty soon you'll learn them all, since there are many similarities except for syntax. I would suggest you learn a 'friendly' language like Java or Python for the first time, it will get your feet wet for later. c++ is really cool though.
Yeah, as a working programmer I'd say that the first programming language you learn doesn't matter too much. Most important thing is that you keep at it :)
Another wonderful video.
This is, hands down, the best "AI consequences" (not only financial) videos I have seen so far. 👏👏👏👏👏