turns out AI can't mimic Philip fully. and it is quite terrible at, what it sort of can do.

Makes him sound like an old British gentleman

A bit like Richard Ayoade sometimes too

As does the AI script. It's near-meaningless drivel, if I wanted to hear something like this I would turn on reality TV.

​@@lollo863oh yes, that's more accurate

AI Philip sounds like Mark from peep show

Yeah haha changes from English posher accent to American haha

Noticeably American, yea lol

As an American, AI Philip has a less noticeable accent. ;)

This statement is false!

Interesting difference in behaviour because I hit like immediately upon watching almost any video. Sometimes I have to go back and revoke the like because the video was bad

@@BonJoviBeatlesLedZep I immediately dislike any video upon watching it. It has to be good enough for me to put the effort into reversing it into a Like. (just kidding)

No Caboosing, must be AI generated video

rip the cabooooooooooooooooooooooooooooooooooooooooosing

huh, the disclaimer shows up for me, too (also mobile). After reading a bit in the link provided in the disclaimer, it seems that Philip put it there by himself, so it's not auto-tagging with some smart'n'fancy KZfaq system

It's not showing up on browser, interestingly enough

the term "synthetic content" is so 2020s it makes my head hurt

ai generated jerma comment

rip off boris johnson

@@-KozloffLMAO

@@SpahGaming damn you caught me

Yeah I'd love to see a sort of retrospective of AI progress on a periodical basis from Philip. It'd be great for future reference.

At this point, 2 years is an eternity for AI development, so I wouldn't put anything to be impossible. Real question is if all this fever dream stuff actually starts replacing jobs on mass I think

@@blenderalien 2 years in AI time is like a decade in normal time

people are going paranoid, they want to sound smart by pointing out a small inconsistency and call out "THATS AI!" But they don't realise how they're actually quite dumb.

it's weird cuz in 2 years it will be indistinguishable and definitely not monotone and get accents right

@@keatonwastakento be fair, they already did that years ago. Like the reliable "that's CGI lol, idiots" comments under any boston dynamics video... but yeah, now we have so much more of it and it's only gonna get worse from now on

probably because there are legit channels with AI voices and AI scripts, its a pandemic

AI will not be good for the schizophrenics

scary cat

I would just add these indicators to a video with no AI at all >:)

I think this is one of the issues with how people approach AI, why do we only need AI indicators? Should films have to say when its an actor on screen doing stunts or a stunt double, or when a scene is filmed on location or using a green screen, or when something is CGI or practical effects? Should any photo which has gone through photoshop, even for things like colour or light balancing need to be flagged as edited? Media has been full of tricks to show you (sometimes) believable things and while some of them, mostly relating to films, you can expect an audience to understand that what they're seeing is to some extent not real, that's not true for all of them. Requiring only AI content to be flagged as not being "real" seems to be an attack on AI rather than a defence against people being misled. I think it's more appropriate for people to maintain a healthy scepticism over the media they consume, like people do when they watch films like Avatar. Requiring only AI content to be flagged or indicated would allow people to pass off material created or modified through other means as being "real", because it would separate out content elements into AI generated and "real".

@@Viking_Cookie Filmmaking? That's a terrible metaphor. Films are immediately recognized as works of fictional entertainment outside of documentaries. Regardless, you seem to have forgotten about the credit reel: Actors, stunt doubles, shoot locations (city and/or studio), and companies responsible for the CGI content are all made clearly visible to the viewer. When I look at KZfaq/Twitter/Twitch content, I'm not informed of whether a human generated it with a string of text, a human created it using Blender, or if an AI generated it by producing its own prompt(s) with not one human involved at all. Is that really acceptable? And this is completely ignoring the issue of accuracy in videos that make statements as facts. An entirely separate can of worms, that one is. "Hey that tidbit of information sounds like bullshit. Let me question the guy who made it. Oh wait, no one made it."

thats interesting, my experience this video has been a lot of listening to what he's saying, then a very jarring moment where either the voice suddenly sounds dead (and American), or the script just sounds totally wrong to what I expect from his writing style (and sure enough look up and the respective element has been AI-ified!). I think the AI used here for script-writing where it takes over is very big on "amplify the previous point with jargon, or roll onto the most obvious progression given the circumstance" while managing to go quite off track or using metaphors that don't seem right. Valid script writing, but it doesn't feel like his work.

hey that’s not true at all, i didn’t hear any gunshots in the background

"Atomic Bombs & Taxes" what a good title for a song, or a band.

@@Aaron067 This comment was written by AI.

Sounds like that one text to speech voice that cusses alot

You mean when it *gained* an American accent?

@@klanko64 no, when people say lost an accent they mean stopped having an accent and started talking like a normal person.

​@@TheManinBlack9054 You're probably (hopefully) just joking but usually "to loose an accent" just means to change from one accent to another. In this case it changed from his usually accent to a strong and weird american accent.

​@@TheManinBlack9054 and what does a normal person sound like?

What is different about this and the prior video.

@@Anon.G The prior video had a more unbelievable AI voice.

What's even worse is that we can't fully trust tools either. By nature, tools will always be behind the latest AI advances. It's really depressing to see a future where you simply won't be able to know for certain if your "is this AI?" doubts are true or not... I don't want to have to put every single bit of information I see through an AI detector either :/

The information era is over, we are deep into the disinformation era.

That's the thing though, if AI really does become indistinguishable, there will literally be no other option except relying on verified journalism. The only and only way to ever tell if anything is real again will be to have a network of professionals with a stake in their careers verify it.

you just wrote "stile" lol

@@daver12591 Ah yes, I corrected it now I think. English isnt my first language and I already need to focus on spelling if I want to write correctly in my mother tongue.

@@indestructiblemadness8531 fair enough

ronronronron

Yep, the exact same manner of speaking and cadence.

I mean you can still do it for fun and for your own enjoyment. You do it because you love to, not because you have to.

@@TheManinBlack9054 it still feels bad that something which would take you hours to create can be made in seconds by a machine

For me it worked well since I lost all interest in programming then AI programming started to become a thing and now I like it again, it makes it way faster and less tedious.

Yeah I've had this feeling 4 times in my life now. I used to play chess a lot but AI beating it made me lose interest in the game. I used to like art and animation but AI has made me lost interest in that 3 years ago. Now I'm also starting to lose interest in programming and electrical engineering because it is also able to be automated by AI. Somehow even though it was enjoyable first the fact that it is now automate-able makes it lose that specific creative appeal it once had. I kinda fear for it because once all creative outlets lose this specific spark it will mean the only engagement humans might have could be consumptive in nature, which is just soulless and would, in any other time be considered dystopian.

Meh, the world ran on misinformation and stupidity already, this won't change anything.

For I welcome our AI overlords.