Is Artificial Intelligence Taking Over? | The Russell Howard Hour

From The Russell Howard Hour, Russell takes a look at the recent rapid rise of AI, and if it's going to take over anytime soon.
Welcome to the official Russell Howard channel! Be sure to subscribe to watch the very best Russell Howard clips!
Follow the show on Twitter at / howardhour
Follow Russell on Twitter at / russellhoward
Become a fan on Facebook at / officialruss. .
SUBSCRIBE NOW. Watch new episodes every Thursday at 10pm on Sky1
#RussellHoward

[@lenkacfk7155]

GIVE THIS MAN A SHOW AGAIN, FFS!!!

[@zakgault4209]

He does? Just not for free.

[@teddieprox2307]

​@@zakgault4209on what channel or service ?

[@missyquill5709]

@@zakgault4209 Where? I will literally pay!

[@Shaun_3114]

Thursdays on Sky 1. I think it's 11pm.

[@missyquill5709]

@@Shaun_3114 What's the show called please?

[@spiritusmundi70]

I'm not worried about artificial intelligence, I'm worried about natural stupidity.

[@russianbot4418]

I've been A-B testing certain types of comments on the YT comment section for a while now. I still can't tell which are just illiterate tools and which are repurposed AI chatbots that can't read text formats that humans should be able to figure out in seconds.

[@centuri6510]

@@russianbot4418Well that’s ironic coming from yourself! 😂

[@russianbot4418]

@@centuri6510 It's Okay, I'm open-source and thus highly adaptable. 😃

[@Kayenne54]

who do you think is inputting data into this monstrosity?

[@spiritusmundi70]

@@Kayenne54 I think you are

[@robertpearson8546]

My two favorites are: 1) Lawyers used ChatGTP to prepare a brief for the court. When the judge checked the cited decisions, he found out that they either did not exist or the decision was grossly misinterpreted. 2) When an Artificial Idiot was used to diagnose skin lesions, it determined the lesion was cancerous when there was a ruler in the picture.

[@abderrahimbenmoussa4359]

AI are there to help, not to decide and are as good as the data they are fed. So dumb Ais are quite often the result of dumb input.

[@michaelwells4802]

The legal industry is literally one of the most vulnerable industries to be replaced with AI. One example of failure does not equate to the huge increase in efficiency that AI has introduced to the legal industry across the board. Lawyers are definitely going to be losing their jobs. If done correctly, AI can replace a whole host of jobs. The entire global economy is going to need a re-work because of it, and people affected by it the most will likely gain a universal income of some sorts due to the taxation of the robots that have replaced their industries. You will basically be gaining an income just for existing.

[@WebToolkit]

The skin cancer AI was 100% default of the developers. If you're going to poison the well of the data set you can't be surprised at the outcome is useless. Every single example of cancer had a ruler in the image and all of the examples of non-cancerous growth didn't have a ruler in the image. There is blame to be laid here but it's not at the feet of the AI it's at the idiot researchers who did think it through.

[@BadWebDiver]

Always remember: Garbage In, Garbage Out - GIGO.

[@jamesmorgan7629]

Russle

[@figplucker908]

That robot feeding the baby. I almost choked laughing. Am I a bad person?

[@denisepurcell4031]

Not at all, unless I am too 😂

[@katherineleblanc9218]

I had to bury my face in my pillow before I woke up the household. You are not alone.

[@Boyakishan]

It’s not, it’s kinda funny

[@figplucker908]

@@Boyakishan I rewound it more than a dozen times, I shared it 20 times and I’m still laughing 3 days later. Convinced there is a special layer waiting for me in hell.

[@Boyakishan]

If that's what you think, sure. But it's just a clip on the internet. It's bad, but it's not what you wanna do to a child sooooo@@figplucker908

[@Vee-jc1qh]

There's definitely no artificial intelligence with Russel Howard. Absolutely brilliant and spot on. I am not a robot!

[@blindmown]

Exactly what a robot would say 🤔

[@Vee-jc1qh]

@@blindmown Malfunction, malfunction.

[@Boyakishan]

@@Vee-jc1qherror

[@jamesmorgan7629]

Russle Howard is a fool

[@Boyakishan]

@@jamesmorgan7629 ???

[@nickschumacher4309]

If robots take out humanity, it'll likely be because someone showed it social media as a learning method.

[@Soul_404.]

Ultron: Saw the internet, decided humans had to die

[@nickschumacher4309]

Honestly, even as a human, one look at social media and I think humanity doesn't deserve to live.

[@weswheel4834]

I thought back to Russell's reference to computers taking things literally/misinterpresting when I read this. "Computer, take out humanity". (later) Computer says: "We went out. I had the fish, humanity ordered the chicken. A lovely evening was had by all."

[@polybius-jy5rc]

i heard that microsoft let an ai learn through twitter and it was, well not crazy, IT WAS A FUCKING RACIST

[@stevenredpath9332]

I believe that they did and the AI went totally fascist. They shut it down.

[@lynnpardo]

A true comedic genius and social realist.

[@lars_larsen]

The more I learn about the development of AI in my studies, the more I get scared of the effects of capitalism on AI development. If AI projects are funded by the people, and the primary motivation behind them is making the world a better place for everyone, I think it'll probably mostly be fine. But, when an AI projects are funded by a few people who are primarily motivated by getting a return on investment, that is when we get misaligned and dangerous AI products released upon the world to turn a profit.

[@dan-nutu]

All of this going on and me? I'm (still) terrified people read "nuclear" as if it was spelled "nucular" (2:34)

[@ezura4760]

Mine is the incessant usage of “should of”…

[@Kayenne54]

High five!!! I've had to delete so many rants about that word, spoken like that, from so many YT videos...If I'd left them I'd be banned off the site. My next most hated mispronounced word is "Acrost" instead of "across". There is NO FREAKING T in that word...sorry. Lost my composure. Hearing people in interviews saying "I axed about that" instead of "asked", "I drugged him out of the house" instead of "I dragged..." And anyone under forty years of age whose every second word is "Like". And "it is what it is" has now made it to No: 5 in my most disliked expressions.

[@Grimfang999]

Theres a streamer that played chess against chatGPT. It was awful at it. AI programs are only designed to do one specific job, so getting an AI designed to mimic human speech and get it to do just about anything else and it will always do it completely wrong. True AI is still very, very fary away.

[@CreamAle]

Keep in mind... What we have now, is as bad as it will ever be. This is the start, not the peak. Also when it comes to single use ai, that's not necessarily true as even gpt has plugins which allows it to interact in other ways. Am I personally worried? As a musician and artist? No. There's already some fantastic ML projects that clones voices almost perfectly and can generate music really damn similar to what I would write. But at the end of it all, these are just tools and tools by themselves aren't evil. How we use them, that's the issue.

[@HellHunter00]

Artificial Super Intelligence(ASI) is what people should be concerned about and what extremely intelligent people like Stephen Hawking have raised serious concerns about(BBC - Stephen Hawking warns artificial intelligence could end mankind - 2 December 2014). If we reach technological singularity, things can very quickly get out of human hands. AI as it is isn't particularly scary but will lead to massive job loss/economic restructuring on a scale that humans have never encountered. Factory level machine automation job loss would pale in comparison. If we'll ever create ASI remains to be seen although IMO it will happen in the not too distant future. Its implications we can never know. It may be the best of times, it may be the worst of times(it could be the age of wisdom, it could be the age of foolishness, sorry, had to).

[@abderrahimbenmoussa4359]

ChatGPT-4 can call on other AIs. So, in a year, if you want to play chess against chatGPT, it will just call on an AlphaGo extension and crush anyone lol That is the thing with AIs can work as agents for each other. Besides, these models were not "programmed" as such. They learned by themselves, so they can develop habilities that were not expected like how language models can in some cases model physics or create music.

[@Blatant88]

When ai is using quantum computers that's where we lose it

[@nobbyfirefly57]

but it does give great cuddles! ^w^

[@CollinMcLean]

The problem with Artificial intelligence is that it can be stupid in ways we can’t predict or understand. It’s the Black Box problem.

[@nerfherder4284]

Yup. The real danger is that we trust without verifying

[@barcelonachair6487]

Similar to the people that its modelling? :)

[@CollinMcLean]

​@@barcelonachair6487 You could say that but unlike a human an AI can't rationalize itself, it's just trying to operate within the parameters of it's programming which adds a new layer of unpredictability.

[@b.b.8955]

The guy yelling at the air freshener thinking it was Alexa probably thought that every time he offended it, it made its displeasure known with knotty pine scented farts directed at his face. 😂

[@abdalleahmed4613]

Howard is the funniest British comedian out today. He should take this show to the states

[@abdalleahmed4613]

@johnjordan3527 what does that have to do with my comment?

[@Domeschannel57]

He’s saying your taste is weak bruh, which it is, the guys being doing the same shtick since Good News when I was like 10, don’t get me wrong he’s entertaining and enjoyable to watch but funniest uk comedian today? Crazy statement

[@kapervee]

OMG That baby robot. 😂 I almost choked in my laughter. 02:10

[@bambieyedgirl7846]

🤣🤣🤣🤣🤣

[@highdownmartin]

Baby feeding robot needs it’s own series

[@albertokari1374]

I love this guy😂😂...much love from Kenya🇰🇪🇰🇪

[@hoebywan]

Trust me, trains in Britain have taken that 30mph limit to heart.

[@wayne3449]

Does anyone even use trains anymore, that's s bad as driving an electric car charged by a generator 😂

[@mastertrams]

@@wayne3449 Uh, yeah, people use trains nowadays. Bear in mind that the rail industry is still running only 60-70% of the pre-Covid timetable (for reasons that are beyond stupid that I shan't get into), the railways are carrying 99% of the pre-Covid passenger numbers. Which means our trains are more full than they were pre-Covid. And especially impressive when we have a notoriously anti-rail government.

[@hoebywan]

@@wayne3449 Nothing is as bad as driving an electric car. Also trains carry so many people the numbers make it an efficient use of power.

[@wayne3449]

@@hoebywan but literally take your independence away 🤦

[@hoebywan]

@@wayne3449 I am independent of the ball and chain that is vehicle ownership. I walk, cycle, bus, train, coach to anywhere I need to be. I pollute the environment less and save more money. You can keep your version of independence.

[@ThePleb]

using AI to talk to deceased loved ones was a black mirror episode a number of years ago

[@internexusx]

The same guy who developed this tech probably saw that Black Mirror episode and say “hey that’s not a bad idea!” However, am I a bad person by thinking that it’ll be hilarious to embarrass my grandkids by calling while they’re using the blowjob robot ?

[@IamtheJayman]

I heard a fun story about AI. A coding camp/class/competition was challenged to create an AI to play Poker. The Winner was just programmed to go All In every time and the more sophisticated AI systems couldn't respond properly to it.

[@itsjustme8947]

Heh, that sounds about like half the people I play against online already! Those fools take all the fun out of playing.

[@itsjustme8947]

Using a multi-million dollar fighter simulator, AI was tested using voice input. The AI would ask if it were performing correctly after each command and it was supposed to be learning how to perform maneuvers literally and instantly. Well, it certainly was taking things quite literally. After ordering the AI to perform a simple left banking turn while holding the g's to a precise 2.2 g's, it asked if it were performing correctly. The pilot commanding the AI said, "That's right. Keep it up." The AI, while in a left bank, immediately and violently pulled the aircraft to the right and straight up.

[@harryallen6051]

nah, I'm not read that shit

[@Boyakishan]

@@harryallen6051An AI took something literally and plane went right up

[@itsjustme8947]

@@harryallen6051 Oh, and because you didn't read about a failed SIMULATOR experiment it didn't happen??? Got it. /smh

[@umwhatamIdoinghere]

​@@itsjustme8947I don't see anywhere in his comment where he claimed it never happened.

[@Hollow_Theorist]

@@harryallen6051 Then don’t read the bloody comment.

[@pb6481]

Foundational large language models can be quite good in getting the nuances in language right, that is something they can be trained on with large amounts of text. However, they are not as good in getting niche facts right. Essentially they generate the next message that is missing in a conversation, so when you're asking questions about some specific fact, you're not querying a database (well, maybe a vector-database, but that's not helping to make a point here). You're asking a question, so it generates an answer (which is what is missing in the conversation), but the facts in that answer are based on probability, and might be completely wrong, but brought verry convicingly.

[@usvalve]

If AI learns from the internet, and the output of AI is published on the internet, eventually AI will be learning from AI. That could be interesting.

[@russianbot4418]

What makes it worse is that over the last 20 years or so many studies of who spends the most time online have been done and the overwhelming data showed that about 60 to 70% of comment traffic comes from those likely from the bottom 30 to 40% of the intellectual bell curve.

[@gazz3867]

It already does. AI generated images are now learning from AI generated images on the internet that were based on AI generated images on the interner, getting increasingly unhinged in the process.

[@Kayenne54]

Yes. We'll end up being only permitted to use one word in any discussion. *"LIKE".*

[@Tulf42]

I'm still worried AI in real life will eventually find and misinterpret all the Sci-Fi films with AI considered the enemy such as Terminator, the Matrix, Age of Ultron etc.

[@Fredy--]

I just love the guy whose like Theres a cat in me house screaming intensly.

[@RuthBhmand]

Do not say the phrase “it’s a fogging pig’ to the AI, you KNOW what it will do with that piglet 😱

[@giordanodsouza9563]

It'll pull a Black Mirror? * Shudders *

[@ashleyspicer4199]

Russell. I've watched every season of the BBC and the sky network. Britain and your tabloids LOL You are a gem!

[@Jovius2k]

I was eating chocolate pudding when the 2 Girls 1 Cup joke showed up. Thanks, Russell! 🤣

[@davidhoward4715]

Okay, I give up. What's "eating a chocolate pudding" a euphemism for?

[@Jovius2k]

@@davidhoward4715 No euphemism, I was enjoying a bowl of chocolate pudding after lunch when I saw that bit of the routine. And if you haven't watched "2 Girls 1 Cup" yet... well, you have been warned.

[@helengorton894]

Yes!!!! Give him a show!!!!!

[@chrissheppard342]

He's been on sky tv for the past 10 years.

[@thehellyousay]

that last guy sounded like seeing the robot dogs scared an impromptu confession out of him

[@flatplatypus]

Brilliant. As usual!

[@sylviacoaker8830]

I love Russel Howard, makes me laugh out loud 😂

[@saoirsevicteoiria2759]

That last man sounded like the most disappointed drill sargent in history.

[@jackcorran4316]

I cannot wait to see this man later this month

[@tundeorosz5937]

Russel!! Please keep it coming! 😂😂

[@CollinMcLean]

We basically created the discount store equivalent of Ultron.

[@vikj1255]

Spot on Russel!

[@1972hermanoben]

The laugh from the audience after that guy had his hed smashed against the countertop 😂😂

[@brll5733]

I do not fear AI. I fear AI controlled by humans.

[@Kakarot64.]

Or at least AI poorly programmed by humans

[@nerfherder4284]

Or the humans that listen to it without thinking

[@Saint_Dan132]

Oh damn x I love our Russell

[@HouseSanity]

Please bring back Russell Howard’s good new it was soooo good

[@adambartlett6277]

Does Russell never age!? Literally looks the same as he did 15 years ago

[@Daktangle]

There's also the plagerism side to these algorithms. The ones used to generated images, I won't call them art as art is human expression, are almost all trained on stolen artwork whether it be images, music, voices, etc.

[@pirobot668beta]

Define 'stolen'. If a human goes to a Gallery, looks but doesn't buy, is that theft? If a human is online and scrolls through an online Gallery, is that theft? Is it only 'theft' if a machine does it...wait a minute...humans set the machine up to peruse a virtual Gallery...are the humans at fault for the 'theft'? If 'stolen' doesn't fit, define 'theft'...

[@JB9000x]

I'm not worried about art as true art is an interaction between two humans: the artist and the observer. AI can create music, a painting, a movie... and it might be entertaining, but it will never replace true art, as when you experience it you will not be moved on a human level to see/hear what another human is communicating

[@willowoates4407]

I got one of those AI scam ads in the middle of this video

[@Youcanttouchmyhandle]

Awesome update from the UK World hugs to the UK

[@tundeorosz5937]

Gotta love him!!

[@soufiannajid3633]

We miss you man ❤

[@user-hd7kk3ks2i]

He’s brilliant 🤩

[@toastyporkpie8540]

how cute: they where taking the hellhound-deathmachines for a litte walk in the park

[@oscargraveland]

@7:35 When the robot dogs walk smoother than their handler (in blue shirt)

[@DTheWhiteWolfD]

Speaking of AI I live in Cambridge and here in some parts of the city we have robots that deliver people’s shopping, they’ll cross the road on they’re own and thank cars and everything

[@NeonNijahn]

Love this

[@RegularBiscuit]

you got this, take your time king

[@Quartz5005]

I would love a comedic movie that just shows an ai robot taking things too literally and is trying to learn what the phrases with double meaning actually means.

[@KorySutherland]

Yet they still can't make editors capable of de-interlacing videos before uploading them to KZbin What a time to be alive

[@complexptsdsurvivor]

I love how Jelly made an appearance on Russell Howard 😅 5:02

[@taunteratwill1787]

Show that AI a photograph of Piers Morgan! 😂

[@puirYorick]

Russell, you've corrupted my image of my perfectly innocent dishwasher now. Thanks, m8.

[@deathybrs]

The frame interlacing is showing in your video capture.

[@a.a.werding2620]

I liked when he had the people using the crosswalk to represent human traffiking😂

[@gabecoolwater4497]

After watching those tiktok videos (or whatever they are), I don't understand how people worry about AI and not humans...

[@mand6377]

😂 thanks for a good laugh

[@scooterdooter]

That explains why Ultron was so messed up; he spent time on Reddit.

[@Jin420]

Omg.. that baby bottle machine.. 😮😂😂

[@GavinColbourne]

0:09 I honestly thought you said w@nking lol

[@chefdean7257]

That baby bottle formula was a knockout.

[@flufwix]

I’m with the last guy! 😂

[@williamtomczak1857]

Knock me up in the morning

[@cameronbrown9080]

You would not want to be related to the last block 😂😂😂😂

[@isabellakeman2016]

there are various negative things russel didn't mention, such as AI being able to imitate voices from recordings which have been used to try and scam people e.g. pretending your daughter is kidnapped. Also adding a naked body to images of fully dressed people

[@katharineeavan9705]

Also stealing artwork, being used to profile people to law enforcement and immigration (we've seen how well that kind of AI profiling goes in job recruitment already), cheat on university assignments (perhaps not a big deal if you're studying fine art, but would you want a doctor who cheated all their assignments?), being used for hacking and spam attacks on websites, taking unsustainable resources to run, being "taught" by underpaid and exploited people in slums and refugee camps... I could go on. The the list is actually quite long. Ultimately Russel is right, the problem isn't that AI is bad, it's that the people making, selling and using it are. Unfortunately, that isn't likely to change any time soon. Billionaires aren't interested in improving quality of life for the rest of us, or providing AI labour for menial and dangerous jobs so the rest of us can be safe and healthy and have more time to be with our loved ones and create art. They're interested in providing AI labour to reduce costs to other employers and remove power from the working classes. Unless there's a truly massive financial pay off to to do otherwise, they want us doing the menial and dangerous jobs while AI spends time with our loved ones and makes art.

[@Nosnibor2023]

Didn't wake up today thinking I'd see Nickocado Avocado in a Russell Howard video...

[@jamesdowman2023]

Same, also was that jelly breaking the keyboard?

[@greyprincebidoof]

Most ambitious crossover since endgame

[@Nosnibor2023]

@@greyprincebidoof nah since the duck billed platypus

[@Roberto-MMF]

is this a new series?

[@RhamanaChan]

Peoples apprehension around AI just makes me think of the old Brasseye clip of a guy walking around holding a placard with the word *TECHNOLOGY* written on it haha.

[@katharineeavan9705]

It's not the technology I'm apprehensive about, it's the motives of the people buying and selling it and the physical cost to us and the planet in terms of non-renewable or environmentally polluting resources and exploited labour. Give me AI that replaces dangerous or menial human labour, or is tasked with improving human quality of life or solving climate issues and I'd be over the moon. It's what most Sci Fi fans have been hoping for for literally a hundred years or more. It's partially what AI was initally conceived of as being for. Hell, even a materially useless AI that's relatively harmless and entertaining would be welcome. Instead, what we get is AI that replaces (or attempts to replace) human artistry for cheaps, more efficiently racially profiles people, and more effectively invades people's privacy and terrorises and harrasses people, all while being built from finite materials mined by exploited workers, being trained by exploited workers and on stolen IP, and running on hardware that's highly resource intensive and damaging to the environment.

[@OrimaJam]

was that Nickocado Avocado and ProGamerJay?! I was not expecting them in russel howard

[@SimbaTvb]

The Black Mirror dogs at the end of the video concerned me

[@peterkarlsson720]

Maybe these Apps will in 50-100 years be smart enough to be called an AI. If ever. It's mainly copy&paste machines today. And some fancy cameras and gyroscopes when it comes to robotics.

[@assilemisawesome]

Freaking hilarious! 😂

[@stevenredpath9332]

There’s a story in I, Robot by Issac Asimov called Little Lost Robot. Guess what the last order it was given.

[@oliversherman2414]

AI is truly halarious at times 😂

[@ripdbtpoo1441]

Reference to 2001? I do hope so...

[@oliversherman2414]

@@ripdbtpoo1441 ?

[@Joronobon]

Fantastic 😂😂😂

[@patrickhaley1312]

Oh yeah, I bet it would have no clue if I were to show it any of my pictures while I was in show choir. Because it learns from the prejudices of humans, it also has their prejudices.

[@lucybarnard3954]

Thought your Nan might call out “Russ where are you” and you reply “I’m coming “ 😂

[@gerardflynn7382]

Artifical intelligence is nothing new. It has been around for centuries. A. I. is more commonly known as Career Politicians.

[@MsAnpassad]

The last letter in A.I. debunks ýour claim.

[@kasondaleigh]

😂

[@serenegeneration]

I think my reaction would be tye same if i saw those things walking the streets tbf lol

[@pirobot668beta]

Simple solution: train a massive AI with nothing but the worst of the Internet. Set it up to serve as a digital 'bad example' database. If you want to test your content for icky stuff, take it to the 'expert' system. We could call it TRUMPd. Transient Rudeness Under Massively Parallel dingleberries.

[@Adzie36]

The satnav was a dashcam

[@Rainbowofthefallen]

Just commenting for the algorithm ❤️🧡

[@sunvie19news]

I love it it’s hilarious but sadly so true .

[@aporue5893]

''too much technology''-mr roboto

[@Generalphoenix8438]

Though ai is here and it's staying. It will and is a part of our lives. But theirs still alot of work needing to refine the programming. If we can get star wars like droids we'll be okay. Similar to the three laws of robotics but we still need to refine the programming so it can keep to the right standards that cannot fully replace humans but needs us as much as we need it. It can be used to help but its the matter of how we restrict and be aware of our actions Its like the magic haddock, ai isn't bad or good but different. We just need to be responsible with it.

[@Penguins247]

I'm pretty certain my GP practice has been run by AI for at least ten years. Either that or the NHS counts death as a clean up.

[@Einomar]

Oh it can guess can it? Wow incredible... 😂

[@nealgrimes4382]

I honestly can't think of anything worse than talking to an A.I version of a Dead it's bad enough talking to the real living ones.

[@anenglishmaninmalaga]

Omg can't stop laughing 😂😂

[@nipplelickle]

0:56 who said from the 90s

[@im_a_shouty_mathew_baynton9867]

so true

[@Rosiedelaroux]

As RUssell got a glass eye

[@sarge420]

Russell is hilarious. Thanks for the laughs, and just a heads up…tRump is buying a flat down the street from you mate.

[@alexshinra6722]

The only real danger is that researcher nail it and they end up repla eing a few more feilds of jobs than planned.

[@francisbrown5469]

If I were Rusty , AI would have to be the scariest ...

[@hardcoregamer264]

honesly we should work alongside AI and not letting it work for us.