CrowdStrike decided to YOLO a bad update to every single customer, all at once, without any testing, or canary releases. Not sure if deliberate action of a disguntled employee, or gross negligence.
100000% Everyone calls AI race a "Gold rush" but people don't look at history and see how devastating the California gold rush was with deforestation, destruction of ecosystems with runoff and by rerouting rivers and mercury mines that continue to contaminate California soil and streams 100+ years later. During the rush only a handful of people got incredibly wealthy and in it's kinda feeling that way right now with zero laws, regulation or consequences for cheating and stealing.... "Allegedly" (in Ellison's voice)
Big time. It's heartening that these conversations are happening on the biggest tech channel when the vast majority of them are mostly ethics-agnostic, or at least fairly parochial when they are.
Ellis's comments at 52:06 hit the nail on the head. "A trillion-dollar company… can drastically alter the entire environment in which we create." The titans of industry want us to think we live in a vacuum where there's no difference between an AI or a human creating something, because it benefits them to do so. There has been a very long and consistent effort to reframe the terms in which all kinds of labour exists, not just art, and this is exactly how we've ended up with garment sweatshops, migrant farm workers, dirty mines (coal in the last century, rare earth minerals in this one), etc. Creative fields really are one of the last realms where the actual human expression of the person making the thing matters, but AI companies are trying to sell us on the idea that it shouldn't.
Human creativity will always have its place. AI will occupy the space where human creativity has already been reduced to a commodity-places where it didn't really matter anyway.
As a father of 2 who's always sick and tired (and current 2xing it with covid), I REALLY appreciate this little reprieve that is the WVFRM Podcast. Thanks guys.
I feel like a lot of the difference between our reaction behind "human-made" and "AI-made" is the context behind the song or art. If a human makes an emotional song, we read into the lyrics and emotion and begin imagining a history of what that person went through which we can try to relate to. When we discover an AI made it, we understand that there is no overarching context for us to relate to or imagine.
I was in Woolworths on Friday afternoon and simultaneously saw all the checkout machines freeze and shut down, I remember thinking wow, that’s weird and interesting, I ended up leaving and not buying because I couldn’t (cash wouldn’t even work) then, when I get home and watched the tv I see it’s a big global crash…
The Cr-48 was intended for testing only, not retail sales. The Cr-48's hardware design broke convention by replacing certain keys with shortcut keys, such as the function keys, and replacing the caps lock key with a dedicated search key
Ellis is a good example of "tell me you're smart without telling me you're smart." How he makes what I think are complex ideas so easily understandable is 💯.
David casually just dropping the coolest dad lore about doing contract work for Samsung
6 месяцев назад+19
Technically the CR-48 was not a consumer device as it was not sold anywhere. It was distributed free to developers and fans who then sold them in a second hand market.
Yeah, a prototype, but the Nexus Q was the first cunsumer hardware launched & made by Google in 2012. It was scapped and basically became Chromecast and launched in 2013.
Love you, Ellis. The progression through the technology era has been so fascinating that we truly aren’t stopping to ask why anymore. Hard conversations need to be had to not allow the human spirit to become a backseat to everything else we CAN, but maybe shouldn’t, do.
Whilst long, I loved the discussion and commentary on the AI segment. Awesome. Would like to see more of that, even if it's in a bonus episode or something
The CR-48 was a prototype device given to a select few (developers, early adopters) as part of a pilot program to gather feedback on the Chrome OS. So technically, the Chromecast was the first consumer hardware product by Google that was commercially available.
For real this part was so interesting! I really hope they do an episode just on AI ethics because every time they bring it up, it is a great conversation, that always seems to get cut short
Complementary information: You actually do not need to train a completely new model for it to “forget” information. There is a process called machine unlearning which can be done with way less costs than completely training a new model that can effectively make the model forget about some topic
C.R. 30a is a County Road that runs alongside a very ritzy bit of coastline in the Florida panhandle. I can totally see this situation occurring there. 23:41
CR51 was a prototype that was seeded to reviewers. Never was sold to the public so it's not "consumer" product. Awesome to go down memory lane about Chromebooks and Chromecasts!
Not sure i like the idea of you guys taking a Meta AI sponsorship, seems like it might introduce a conflict of interest. Why not just take phone sponsorships next?
The cooling fan grip for the R5 Mark II isn’t $750, it’s $399. Also, you can still use older LP-E6 batteries like the N & NH, just not the super old base E6 ones.
Absolutely great Podcast and love what Ellis said so agree about everything he said regarding AI and Humans well done.Now on totally different subject.I love the watch band with leather in the Merch MKB and see for the larger one is selling great.Also the Shirt you are wearing very nice and so is Andrew's.Thanks so much Guy's really enjoyed.💯✌️
One comment regarding Apple using the Pile dataset (by EleutherAI). In software development, it's a common practice to use open-source libraries or components in your products. However, if there is a bug caused by that or, even worse, a data breach caused by the security bug in the open-source library, it's not the library to be blamed - you are fully responsible. I'm not a lawyer, but if it worked the same way here, I would say it should now be Apple's problem to deal with.
That photo of the truck is real, its from 3 years ago, or more. The frame is powder coated red and it has a major lift kit. I've seen all kinds of trucks like that over the years, especially at SEMA show (one of the world's biggest car shows for industry). I've seen every colour you can imagine for frame colours. They had an aftermarket metal bumper it looks like, and the bumper does look like it was damaged in the accidebt. Its hard to make out detail super well since the photos is a bit grainy.
The R5 MkII is perfect on paper if you’re already invested in the RF mount. The problem is it’s nearly identical to the Nikon Z8 but $300 more expensive (not counting that the Z8 is usually on sale). Canon’s lens ecosystem is also more restrictive.
I love Ellis’ ethics mic-drop. Marques’ devils advocation this episode made me feel a bit weird… I know his intention was to create conversation, but idk it just feels icky that anybody, and many will, have that stance on AI generated art. And David saying, “but also, I think we just need to care more about the human stuff” is so true. Art = human experience! Maybe the day AI becomes sentient I’ll change my tune, but for now, nope!
The main difference in computer inspiration and human is that humans feel the music they listen to and want to express that to listeners looking for them to feel a new/certain way. When computer does it the outcome is better but loses the “feeling” and can become a quick way to make money rather than an artist desiring to express themselves.
1:24:46 quick correction for you guys, old LP-e6 batteries work but they may not last as long. The only thing it doesn't "work" with is the new grip with the cooling fan, in that it doesn't let the fan work but it still powers the camera
I don’t know if anyone mentioned it in the comments but “ fairly” a group of researcher funded by the French government proved that it’s possible to train a large manage model without infringing on copyright or the need for permissions.
I’m sure there were more data scientists here who share similar points, but feels like you do not see the scale and complexity of the datasets you talk about. The process of creating dataset is simple: 1. You need data to train model. This data should contain specific information and properties. For example photos of cats vs photos of other animals. 2. You hire team to go outside and take 20 thousand pictures of cats and 20 thousand pictures of other animals. 3. You train model and hope data was good enough. Or you start over. Question: With all money of the world how much time do you need to take 40 thousand photos? From business perspective it almost impossible. This is why Google trains models using CAPCHA. This is why you allow to process data in cloud in Apple t&c. You just need to try to build a decent model to classify something (super simple) but not using public datasets and without scraping web.
I had the Canon EOS-5 back in the day and that was so dope. It only had 3 focus points in a horizontal row tho. It worked pretty well for something that came out during the late 90's.
It was so funny for me, Marques building up tension by closing some phones and put aside before start talking about the Apple Home Pod color haha 😆😆😆😆 Top acting
I do not want eye tracking. I use an r5 and a c70 regularly. I am looking at so many different elements on the frame when I am evaluating what I want to shoot, I am not just focusing my eye on the focus point. I am checking composition, background blur, color edges, and more. Unless the eye tracking only focused on the one point then locked while I evaluated the rest of the photo before clicking the shutter. This all takes place in the span of sometimes a just a few seconds, but to only focus my eye on the point i want in focus means I have to ignore the rest of the photo until after the shutter is clicked which sounds absolutely unacceptably terrible.
Even betterhelp would be a more ethical sponsor then meta, and thats saying something. Remember when Marques did a puff interview for Mark Zuckerberg AFTER he caused the deaths of thousands in Myanmar.
Another great episode , AI using RUclips to train itself!!! And definitely can't wait for next week's podcast on Microsoft and CrowdStrike outage 😊❤❤❤.
It’s videos like this that I truly understand how Marques is where he is. For him to be visibly upset about the whole RUclips data scraping thing, but still play devils advocate in order to entertain us and inform us is amazing.
When LEGO changed their plastic compounds to be more renewable, the colors of all the bricks changed. VERY SLIGHTLY, but its visible with a dorect aide by side comparison. Its possible the Homepod is similar
Marques's question about humans versus robots is brilliant. The difference is not just moral, but also business differences. When a human watches a video on RUclips, RUclips sells an ad to that human, or a human is paying for RUclips Prime, while you can't sell an ad to a robot. So robot needs to pay for the data.
The first Chromebooks for sale were manufactured by Acer and Samsung - they were announced at Google I/O in May 2011. The Chromecast was the first product they made for consumers :)
are you guys considering posting the video format on Spotify? im not sure what all goes into that but a few other pods i follow have been adding their videos to the site/app
The Chromebook in question, was manufactured by Acer in 2011. It was also accompanied by a Samsung Chromebook, both of which were announced in a Google press conference that year. Google bought RUclips in 2006. And then went on to design something you could watch it on, The Chromecast, which was released in 2013. The Chromecast was designed by Google but it was manufactured by "Outside contractors". Google does not disclose who these contractors are, but has stated they were based in Asia. David gets the point.
When this video podcast chanel reaches 1 million. Ellis and Adam you need to make Marques, Andrew and David have to do a live podcast. Doesn't have to be on Wednesday recording. And then you edit afterwards.
A written summary does not replace the video. A summary does not give you every detail if it's not retelling minute for minute what is happening in the video. Its similar to news. News does not share an opinion but summarises facts. The excact wording is copyrighted but the facts aren't. Stating MKHBDs opinion is allowed.
OK. So at 50:30, there's a scenario where a double blind test is proposed where people use the same song to show to people and say "AI made this" or "Human made this" ... Except AI didn't make it, it just spliced ten billion songs together to make a new one in its black box. Maybe some would argue that IS creating in a nutshell, but I would argue it's just automating the programming of a song.
The Chromebook Cr-48 was for testing and not for retail sale. Partners like Acer would release their versions. The G1 was HTC. The Chromebook Pixel did come out in February 2013 while the Chromecast came out July 24th, 2013. David should get the point, while Marques was close he was wrong and Andrew would have got a point if Marques had but he also was to vague need to be more specific.
according to google fandom wiki, the CR-48 was made by Inventec, not Google. It also was built to test the first version of Chrome OS and wasnt intended for retail sales (ie. not a consumer product)
No body is pointing a gun at Apple's head telling them to train an AI model, If they can't find legal datasets They can simply build one (I'm sure they can afford to) and release some weird commercial pointing out that the only reason OpenAI and Google are ahead is because they're training their models on stolen user data, it could actually pressure them and make them do things the right way.
Hey all! Could you please bring Brian Greene to the podcast and talk to him (Read: listen to him) about how the fabric of space is woven with wormholes? It would blow all your minds collectively.
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
Can't wait for next week's podcast talking about WTF happened to Microsoft and Crowdstrike
Why they don't make live-stream format podcasts?!
CrowdStrike decided to YOLO a bad update to every single customer, all at once, without any testing, or canary releases. Not sure if deliberate action of a disguntled employee, or gross negligence.
Yuppp
My work was going bonkers because of this
I sold my position of CRWD at 9:30 AM market open. In that business you can break trust only once, gg.
the advertisements are going crazy recently, timestamps are a gift from god
I was also like wow, I pay for premium but this is a lot
ReVanced, NewPipe, etc. Everyone should be using these so RUclips gets the message.
@@AirBnBashOfficialhow are you getting advertisements if you have premium?
@@afgh1408maybe they mean the sponsored parts?
@@corvacopia you are probably right
I am deaf and truly appreciate the accurare captions. Thank you for covering that.
Sorry friend. I cannot read sign language. Hope you said something nice. Good day.
Ellis always coming in clutch with the slam dunk on topics regarding ethics or responsibility
100000%
Everyone calls AI race a "Gold rush" but people don't look at history and see how devastating the California gold rush was with deforestation, destruction of ecosystems with runoff and by rerouting rivers and mercury mines that continue to contaminate California soil and streams 100+ years later. During the rush only a handful of people got incredibly wealthy and in it's kinda feeling that way right now with zero laws, regulation or consequences for cheating and stealing....
"Allegedly" (in Ellison's voice)
For real Ellis can preach 👏
10000% agreed
Big time. It's heartening that these conversations are happening on the biggest tech channel when the vast majority of them are mostly ethics-agnostic, or at least fairly parochial when they are.
Yet they had no response to ethics or responsibility when Louis questioned it about their Apple facility visit.
Ellis need his own show
he got aura
@@slick3996 So much Aura
Ellis's comments at 52:06 hit the nail on the head. "A trillion-dollar company… can drastically alter the entire environment in which we create." The titans of industry want us to think we live in a vacuum where there's no difference between an AI or a human creating something, because it benefits them to do so. There has been a very long and consistent effort to reframe the terms in which all kinds of labour exists, not just art, and this is exactly how we've ended up with garment sweatshops, migrant farm workers, dirty mines (coal in the last century, rare earth minerals in this one), etc. Creative fields really are one of the last realms where the actual human expression of the person making the thing matters, but AI companies are trying to sell us on the idea that it shouldn't.
👏 well said
Human creativity will always have its place. AI will occupy the space where human creativity has already been reduced to a commodity-places where it didn't really matter anyway.
@@Shotblur bullshit. Which gen AI company do you work for?
As a father of 2 who's always sick and tired (and current 2xing it with covid), I REALLY appreciate this little reprieve that is the WVFRM Podcast. Thanks guys.
Formally requesting a 2 minute Ellis rant every episode
the way ellis raised his mug as he said its time to recognise the bean XDXDXD
It's more of a huge grain of rice, or yes, a pill
All hail! 🙌🙌🙌 🫘
Ellis singing “BBL DRIZZY” at 51:32 😂😂😂
I feel like a lot of the difference between our reaction behind "human-made" and "AI-made" is the context behind the song or art. If a human makes an emotional song, we read into the lyrics and emotion and begin imagining a history of what that person went through which we can try to relate to. When we discover an AI made it, we understand that there is no overarching context for us to relate to or imagine.
Listening to that large segment and then going straight to a Meta AI advertisement is wild
We need a special episode tomorrow about Microsoft and Crowdstrike, this is huge!
Very interested in what Crowdstrike did to have this big of a mess up.
I was in Woolworths on Friday afternoon and simultaneously saw all the checkout machines freeze and shut down, I remember thinking wow, that’s weird and interesting, I ended up leaving and not buying because I couldn’t (cash wouldn’t even work) then, when I get home and watched the tv I see it’s a big global crash…
The Cr-48 was intended for testing only, not retail sales. The Cr-48's hardware design broke convention by replacing certain keys with shortcut keys, such as the function keys, and replacing the caps lock key with a dedicated search key
The Chromebook CR-48 was a prototype. I don't think it was ever commercially available
Get Ellis screaming "ALLEGEDLY!!!" On the sound board.
Ellis is a good example of "tell me you're smart without telling me you're smart." How he makes what I think are complex ideas so easily understandable is 💯.
The "GYATT" at 1:23:31 needs to be on the soundboard
You're so right
David casually just dropping the coolest dad lore about doing contract work for Samsung
Technically the CR-48 was not a consumer device as it was not sold anywhere. It was distributed free to developers and fans who then sold them in a second hand market.
How does bro not have a username? 🤯
@@just_mdd4Probably put an ansii space character for a name
And now my name shows in sub-comments !!! ¯\_(ツ)_/
The Chromebook CR-48 was released for testing only, not retail sales.
Yeah, a prototype, but the Nexus Q was the first cunsumer hardware launched & made by Google in 2012. It was scapped and basically became Chromecast and launched in 2013.
I don't think the Q ever hit the retail market. I think the only people able to get one were the attendees at IO 2012
What about Chromebook pixel? I read on internet that was launched before Chromecast in same year
1 x Intro
2 x Trivia
4 x Stories
6 x Sponsors segments
Quiet week huh!
😭 tbf it's a fraction of the total vid length
still wild though
can’t possibly bother you that much just skip the chapters it literally takes 1 second
Heaven forbid content creators get paid for making quality content
SponsorBlock is a great browser extension
This is a relaxing podcast to listen to while taking a nap.
Love you, Ellis. The progression through the technology era has been so fascinating that we truly aren’t stopping to ask why anymore. Hard conversations need to be had to not allow the human spirit to become a backseat to everything else we CAN, but maybe shouldn’t, do.
Whilst long, I loved the discussion and commentary on the AI segment. Awesome. Would like to see more of that, even if it's in a bonus episode or something
Easily, one of the best & most nuanced critiques of AI I’ve heard so far. This was great!
The CR-48 was a prototype device given to a select few (developers, early adopters) as part of a pilot program to gather feedback on the Chrome OS. So technically, the Chromecast was the first consumer hardware product by Google that was commercially available.
Love everything that Ellis said!!! 👏🏽👏🏽 We need more of Ellis!!!
Amazing discussion at 46:35 until the trivia break! Possibly the best discussion I've seen on the topic. "...more at eleven!"!
For real this part was so interesting! I really hope they do an episode just on AI ethics because every time they bring it up, it is a great conversation, that always seems to get cut short
Back home from Greece! 🇬🇷 and it’s time for the best Podcast, after the best holiday! Thanks for what you do guys ❤️
For what it’s worth I enjoyed the longer podcast this week. I feel 90 minutes is good 👍🏻
"Beans come in all shapes and sizes" is what I needed after today's outage. TY Ellis
So far this is my favorite episode. Really gets you thinking.
Complementary information:
You actually do not need to train a completely new model for it to “forget” information. There is a process called machine unlearning which can be done with way less costs than completely training a new model that can effectively make the model forget about some topic
C.R. 30a is a County Road that runs alongside a very ritzy bit of coastline in the Florida panhandle. I can totally see this situation occurring there. 23:41
I love the longer episodes. I wish they would all be 3hrs long
Ellis’ take on supporting people is awesome! 👏🏻
Andrew is the best. Always remembering you are in a podcast and explaining stuff to people listening and not watching :)
47:59 hear, hear ellis!!!!
CR51 was a prototype that was seeded to reviewers. Never was sold to the public so it's not "consumer" product. Awesome to go down memory lane about Chromebooks and Chromecasts!
Guys bring back the 'ive been doing x for a week now' jokes.
"I've been using Linux Mint for a week now, here's my thoughts."
The CR-48 was a developer device
Not sure i like the idea of you guys taking a Meta AI sponsorship, seems like it might introduce a conflict of interest. Why not just take phone sponsorships next?
Nobody is paying attention to this. I think it's just up to us to trust in mkbhd.
It’s not that deep
@@Roxas1711it kind of is dude.
I am not too sure. However, I believe it's Vox who controls the sponsorships for the podcast.
What does a Meta AI sponsorship have to do with the waveform podcast?
The cooling fan grip for the R5 Mark II isn’t $750, it’s $399. Also, you can still use older LP-E6 batteries like the N & NH, just not the super old base E6 ones.
Absolutely great Podcast and love what Ellis said so agree about everything he said regarding AI and Humans well done.Now on totally different subject.I love the watch band with leather in the Merch MKB and see for the larger one is selling great.Also the Shirt you are wearing very nice and so is Andrew's.Thanks so much Guy's really enjoyed.💯✌️
One comment regarding Apple using the Pile dataset (by EleutherAI). In software development, it's a common practice to use open-source libraries or components in your products. However, if there is a bug caused by that or, even worse, a data breach caused by the security bug in the open-source library, it's not the library to be blamed - you are fully responsible. I'm not a lawyer, but if it worked the same way here, I would say it should now be Apple's problem to deal with.
“human beings matter more than computers!” ❤️
Marques good job in Mr. Beast video 👏🏻
He was like, the only one that I noticed I watched so I was upset when he got out. Lol
Shout Out Camp Fiesta! Proud of the fans. Y'all are a great bunch
That photo of the truck is real, its from 3 years ago, or more. The frame is powder coated red and it has a major lift kit. I've seen all kinds of trucks like that over the years, especially at SEMA show (one of the world's biggest car shows for industry). I've seen every colour you can imagine for frame colours. They had an aftermarket metal bumper it looks like, and the bumper does look like it was damaged in the accidebt. Its hard to make out detail super well since the photos is a bit grainy.
The R5 MkII is perfect on paper if you’re already invested in the RF mount. The problem is it’s nearly identical to the Nikon Z8 but $300 more expensive (not counting that the Z8 is usually on sale). Canon’s lens ecosystem is also more restrictive.
CR in the truck photo could mean County Road. Often seen here in Texas (where lifted trucks are found often)
Yeah that picture is not ai. The red underneath is the suspension powdercoat
Yes, horizontal stoplights are common in Texas but rare (or non-existent) in Cali.
1:23:29 never in my life i wound hear Marques Brownlee said "gyatt"
Good catch 😁
I love Ellis’ ethics mic-drop. Marques’ devils advocation this episode made me feel a bit weird… I know his intention was to create conversation, but idk it just feels icky that anybody, and many will, have that stance on AI generated art. And David saying, “but also, I think we just need to care more about the human stuff” is so true. Art = human experience! Maybe the day AI becomes sentient I’ll change my tune, but for now, nope!
When are you talking about Louis Rossman's video?
lol @ David wearing the same T-shirt both days….
*Been there, brother.* 😂
The main difference in computer inspiration and human is that humans feel the music they listen to and want to express that to listeners looking for them to feel a new/certain way. When computer does it the outcome is better but loses the “feeling” and can become a quick way to make money rather than an artist desiring to express themselves.
Pays for premium to avoid ads, creators start including ads in the content. Fml
Bruh I said the same ish to myself the other day. It really pisses me off.
Just skip them? There’s plenty of extensions and ways to easy skip them
Yeah that’s definitely a thing. SponsorBlock is a Google extension that does exactly that. It relies on user-input timestamps.
@AlbieSkywalker timestamps in the description or if youre manually forwarding there’s a green bar at the bottom
@AlbieSkywalkerYes...there are extensions for that. Just Google em
Was so excited for this HomePod Mini update! Went out and bought 10 if them. Haven’t seen anything this innovating since the HomePod update
1:24:46 quick correction for you guys, old LP-e6 batteries work but they may not last as long. The only thing it doesn't "work" with is the new grip with the cooling fan, in that it doesn't let the fan work but it still powers the camera
What a great conversation. The AI exchange is priceless.
I don’t know if anyone mentioned it in the comments but “ fairly” a group of researcher funded by the French government proved that it’s possible to train a large manage model without infringing on copyright or the need for permissions.
Not really sure about how the certification process is done but my point still stands, if there was a will there would be a way.
I’m sure there were more data scientists here who share similar points, but feels like you do not see the scale and complexity of the datasets you talk about.
The process of creating dataset is simple:
1. You need data to train model. This data should contain specific information and properties. For example photos of cats vs photos of other animals.
2. You hire team to go outside and take 20 thousand pictures of cats and 20 thousand pictures of other animals.
3. You train model and hope data was good enough. Or you start over.
Question: With all money of the world how much time do you need to take 40 thousand photos? From business perspective it almost impossible. This is why Google trains models using CAPCHA. This is why you allow to process data in cloud in Apple t&c.
You just need to try to build a decent model to classify something (super simple) but not using public datasets and without scraping web.
I had the Canon EOS-5 back in the day and that was so dope. It only had 3 focus points in a horizontal row tho. It worked pretty well for something that came out during the late 90's.
It was so funny for me, Marques building up tension by closing some phones and put aside before start talking about the Apple Home Pod color haha 😆😆😆😆 Top acting
I do not want eye tracking. I use an r5 and a c70 regularly. I am looking at so many different elements on the frame when I am evaluating what I want to shoot, I am not just focusing my eye on the focus point. I am checking composition, background blur, color edges, and more. Unless the eye tracking only focused on the one point then locked while I evaluated the rest of the photo before clicking the shutter. This all takes place in the span of sometimes a just a few seconds, but to only focus my eye on the point i want in focus means I have to ignore the rest of the photo until after the shutter is clicked which sounds absolutely unacceptably terrible.
Please stop the meta ads
Even betterhelp would be a more ethical sponsor then meta, and thats saying something. Remember when Marques did a puff interview for Mark Zuckerberg AFTER he caused the deaths of thousands in Myanmar.
I agree they’re big cringe. But you know these are their most profitable ads ever, so ya can’t hate!
Nah they deserve the add revenue, shouldn’t hate on it
let them take that bag home. Just skip them, no different from midrolls
I have no issues with it. Get that bag! Lol
damn the first couple minutes made me think they were gonna talk about Louis rossman’s comments on the Apple repair stuff
Same
Another great episode , AI using RUclips to train itself!!!
And definitely can't wait for next week's podcast on Microsoft and CrowdStrike outage 😊❤❤❤.
DAVID IS BIXBY?!???? Plot twist of the century.
It’s videos like this that I truly understand how Marques is where he is. For him to be visibly upset about the whole RUclips data scraping thing, but still play devils advocate in order to entertain us and inform us is amazing.
The Pixel’s new camera bump looks like a webcam 🤣
Good job Ellis on that first trivia question
@Ellis. Marques' "We Like" and "Gyaatt" needs to be on the sound board.
When LEGO changed their plastic compounds to be more renewable, the colors of all the bricks changed. VERY SLIGHTLY, but its visible with a dorect aide by side comparison. Its possible the Homepod is similar
We should vote on the best Pixel camera bump/visor/etc
David not changing his shirt is my favorite thing this week.
Did google ever do something with the temperature sensor on the pixel?
Marques's question about humans versus robots is brilliant. The difference is not just moral, but also business differences. When a human watches a video on RUclips, RUclips sells an ad to that human, or a human is paying for RUclips Prime, while you can't sell an ad to a robot. So robot needs to pay for the data.
The first Chromebooks for sale were manufactured by Acer and Samsung - they were announced at Google I/O in May 2011. The Chromecast was the first product they made for consumers :)
Good point with the museum art, Ellis!
The Meta AI ad is the weirdest ever. Never even heard of Meta AI, literally anywhere ever...
CR-30a stands for "county road" 30a. CR and FR are commonly used road designations in the rural parts of the US
are you guys considering posting the video format on Spotify? im not sure what all goes into that but a few other pods i follow have been adding their videos to the site/app
The Chromebook in question, was manufactured by Acer in 2011. It was also accompanied by a Samsung Chromebook, both of which were announced in a Google press conference that year. Google bought RUclips in 2006. And then went on to design something you could watch it on, The Chromecast, which was released in 2013. The Chromecast was designed by Google but it was manufactured by "Outside contractors". Google does not disclose who these contractors are, but has stated they were based in Asia. David gets the point.
43:01 Thanks to David I set my Timers, Alerts and Reminders. 😉😄
46:30 - 48:00, Thank you Ellis for talking some sense.
When this video podcast chanel reaches 1 million.
Ellis and Adam you need to make Marques, Andrew and David have to do a live podcast. Doesn't have to be on Wednesday recording. And then you edit afterwards.
A written summary does not replace the video. A summary does not give you every detail if it's not retelling minute for minute what is happening in the video. Its similar to news. News does not share an opinion but summarises facts. The excact wording is copyrighted but the facts aren't. Stating MKHBDs opinion is allowed.
David cooked with the title
Amazing episode guys! I wonder what apple’s response will be.
OK. So at 50:30, there's a scenario where a double blind test is proposed where people use the same song to show to people and say "AI made this" or "Human made this" ... Except AI didn't make it, it just spliced ten billion songs together to make a new one in its black box. Maybe some would argue that IS creating in a nutshell, but I would argue it's just automating the programming of a song.
The after credits are always gold!
The Chromebook Cr-48 was for testing and not for retail sale. Partners like Acer would release their versions. The G1 was HTC. The Chromebook Pixel did come out in February 2013 while the Chromecast came out July 24th, 2013. David should get the point, while Marques was close he was wrong and Andrew would have got a point if Marques had but he also was to vague need to be more specific.
according to google fandom wiki, the CR-48 was made by Inventec, not Google. It also was built to test the first version of Chrome OS and wasnt intended for retail sales (ie. not a consumer product)
"You can't just.. let people vote" Iconic David line.
No body is pointing a gun at Apple's head telling them to train an AI model, If they can't find legal datasets They can simply build one (I'm sure they can afford to) and release some weird commercial pointing out that the only reason OpenAI and Google are ahead is because they're training their models on stolen user data, it could actually pressure them and make them do things the right way.
Seems the truck hits Porsche pic is over 2 years old from CR 30A near Panama City, FL
Audio listener here. I was so confused what bean everyone was talking about until I flipped over my 8a and I was like ooooh that bean
Hey all! Could you please bring Brian Greene to the podcast and talk to him (Read: listen to him) about how the fabric of space is woven with wormholes? It would blow all your minds collectively.