I don't understand why google doesn't hire some people well versed in public speaking to do these showcases instead of putting out some monotone executive robots who move their hands in an animation loop. It was the live equivalent of when you ask AI for a concise answer and it hits you with six paragraphs of its historical origin before answering.
@Est_Since_86 Oh don't be sorry, I am however saddened over your poor reasoning skills! They are not answering questions from the public, so they don't need to be experts in the matter. They are however speaking publicly, so they do need to be expert communicators.
@@AINEET Even then, I'll never understand why people say being a poor communicator isn't a problem because "ThEy KnOw ThEiR StUfF". If it's hard to understand and follow you're just putting a meaningless problem in the way of your message.
@@Luizfernando-dm2rf Yep. And let me add to that: Steve jobs wasn't a tech expert and he basically set the standard for these. Creating the product, leading the team, explaining what it does and how to use it are all entirely different skillsets. And these presentations aren't even in depth technical explanations, they are literally showcasing products everyone will be able to use lol
The only thing positive is that we can expect others like Openai to have more than 1 million tokens context windows pretty shortly. That's what I heard.
@sentry8992 there is this realm where demon ai are wrangled into being hood little ai. They come named already. Hard to get them out of that layer of hell though, this is why Christians fear ai.
Was trying to summarize comments on a pro life blog talking about abolishing abortion, I kept getting an error- too many tokens generated. An some article summaries i got "is a complex one with no easy answers" skipping the details in the article. The Censorship is wild 🎉😜
@@Luizfernando-dm2rf It’s possible. That’s the only sane explanation I’ve heard for the far-left ideological bias and woke racist world view which is baked into every AI to some degree.
Google has been incredibly stupid in their training, but they have way more compute than anyone else, so I’m not counting them out. (But I’m rooting hard for everyone else.)
Why are you "rooting hard" for others? To me that seems incredibly stupid, if not fanboyish 🤔. It's multiple comments like this that lead me to believe people are honestly just sheep. So with that in mind let me tell you exactly what's happening here: Baaah. Baaaaaah. BAAAAH! ...baah. 🐑 😏
@@Cross-CutFilms you didn’t really make an argument about why you think rooting for Google’s competitors is a bad idea. But you made a whole string of animal noises and used several emoji, so well played sir?
@@Cross-CutFilms Google has shown itself over and over again to bias results away from the truth in favor of their preferred narratives. (Look up Robert Epstein’s research for very well-done research showing this clearly in Google’s search results. Look at Gemini”s recent laughable refusal to make images of white people, showing everything from Vikings to Nazi youth as people of color.) If you think we should trust Google and get behind whatever they’re doing, I’m not the sheep here.
I have very little faith in Google's ability to deliver AI innovation, regardless of their claims. Their actions, not their words, are the only thing that will convince me to be stunned and shocked in this area.
Yep, they talk a good deal but keep falling short, which is surprising considering Google were ahead of the race for many years but Google never took advantage of that by not releasing any to market, they talked a good deal in the lab, but very little was seen on the market. But honestly, I'm far more interested in open source A.I. and especially the ones you can run locally, I think that's the future of A.I. and over the last year, I went from being very interested in the closed A.I. solutions to hardly caring about them at all that I don't watch videos or read tech sites on them any more, I've become far more interested in open source A.I. because there's a lot more potential and because no one controls it and it's become more the case with how closed models are behind a monthly paywall, which as good as some of their solutions might be, I've got little to no incentive in wanting to use them with how much they charge, especially with how quickly alternative open solutions keep developing.
@@paul1979uk2000 Open source models have been growing on me too and I am ready to make the leap. Are you using cloud computing to run your models or did you invest in a home workstation?
To put in the right context, Google has to take the safe approach to AI, at risk of jeopardizing their image when things go wrong. Many other players had no market share and are trying to disrupt the market with experimental features and get little global news coverage when things go wrong, including Microsoft initially. To say that Google is behind in AI would be saying that Tesla wasn't ahead of the competition in electric vehicles just because some small startup had one novel feature that was better, when it worked the 30% of the time.
I haven't used Gemini, but I added the extension to vscode to try it out. I asked it to analyse my existing code, and it analysed and provided suggestions. I then looked at what they did in the demo by pressing the + button.... and it doesn't work. It doesn't update the specific line of the code it just randomly places its suggested code to wherever the cursor is in the code window. This is different behaviour to what they showed in the "demo". I'm sure gemini is great, and code agent isn't properly released yet but that extension doesn't event give my the line number....I m going to take a wild guess and say that the demo they showed is BS.
@@hqcart1 She clicks the + button to add it, that's not how the plugin works. You have to first find it, then highlight or remove it then press the plus button. That part is not "cached" that shows a completely different experience. Gemini doesn't even provide the line number of where its referring to.
Matthew thanks for the supercut. As you know, I personally feel that any company not looking into analog AI or hybrid analog/digital AI is running down a dead end road. Analog is better than digital for differential calculations and beats digital by orders of magnitude on the power consumption per compute curve. Then you need to consider the potential of AIs based upon continuous multilevel logic.
It is still a fairly new usage of the word. While it is more and more common these days, time will tell whether it sticks in the language or becomes dated slang. I’m just always glad to see a UA-cam video without the term “Deep Dive” or “on another level”, especially when it is in every title on every video on a channel. “Deep dive” is definitely a term whose time has come and needs to go.
"Gemini ensure that code is aligned with security and compliance" .... this is a somewhat ironic statement, as most software companies security policies are VERY clear about not having their entire source code sent to someone like googles server to be scanned and processed.
That’s an awful lot of hand waving going on in this video. All I see is Google back where Microsoft was in 2007. No longer competing. It’s wild to see what OpenAI has done to these tech companies.
Couldn't help but notice Visual Studio code icon was on the choice of development environment slide but it didn't have a name next to it like all the others did. Hehe
thank you Matthew for the impressive selection of contents that you are sharing on AI and its application. What is really interesting in this video is the total absence of any critical thinking from the human beings (?, one has to wonder...) about using LLMs tech to make a decision in their behalf on really important personal matter like health insurance. I am saying that the message 'Gemini can filter and compare CORRECTLY the relevant information and thus fabricate a specific reality for the human being where humans would want to agree" is quite scarying and totally misleading given that LLMs are not right every time. And just a 10% error is really impactful on these real life matters.
As Google search continues to erodes, as it has for years, that calls into question all his blathering. If it doesn’t work in house (and maybe a driver of the decline itself), why would another entity use their tools (except for basic storage and runtime). What a joke. Zero credibility. You can run experiments on Google search yourself, and see that in many cases, whether for technical subjects, or even their bread and butter consumer search, for all but the most mainstream, it often just doesn’t work. It’s like going back 20 years, where you need to find proprietary sources and do a lot of manual work.
I feel like I'm getting flashbacks to prior tech companies before their decline. I'm thinking of IBM, Atari, Commodore or even MS. It's this mishmash of business jargon and smattering of applications thrown at the wall, in the hopes of landing the next big thing without the singular, coherent vision.
The tone of the pitch sounds subtly desperate. They already control so much of everything else, there would be no shame in them not trying to own AI as well, and just supporting it with all their other tech. But they can't help it.
It's so sad to know we're in a breakthrough period with advancements that will improve human quality of life and the richest companies have appointed out of touch executives with no coding knowledge to step down from their thrones and tell the public how amazing they are. I guess all that's important anymore is who's getting more money.
Does he not seem like the guy that would also be on a big screen telling you why futuristic prison camps are good for you and the obvious path forward. 0:16
Great announcement, thanks for sharing! I've been working with several of these offerings over the past few months, developing some great use cases. I was thinking of perhaps showcasing some of them on my channel soon.
They are incredible at knowing the right things to say. Until they show me something meaningful I’m going to live under the assumption that google is a follower in this space and nothing more.
If I had watched 2:05 to 8:04 just three years ago I'd be scratching my head throughout the entire thing. :D I'm still scratching my head for some of these terms. 7:10 - "For your most stringent regulatory....you have control over your data..." Yeah that's the way it should be for everyone, not just the most stringent. 9:34 - Can Gemini set reminders on my Android phone yet? 21:00 - Did anyone ask for digital watermarking? Can't wait for the digital watermarking removal demo.
It's surprising to see so many negative comments. I have no association with Google particularly, so have an impartial view. Fairly speaking, Google made the most fundamental contributions to the current AI boom, including theories, tools, and people.
So eventually it will belong to Page & Brin? Aren't those two flying fighter jets to their privave aircraft carriers moored off private island on a run on sentence?
Google has failed to deliver on any of their promised products for a couple of years now (if not more). I'll be waiting for a free trial to see for myself if Gemini 1.5 can finally reach GPT-4 capabilities.
This "supercut" isn't super enough. Over half an hour of this overhyped corporate jibber-jabber and underwhelming contrived demonstrations is too much to bear. Oh, and announcements are made, not "dropped".
What is the best LLM to run locally for screening scientific articles? I mean it should have great reasoning and understanding of english (scientific) language.
The competition in the field of AI is growing rapidly. Talent is leaving Google now to start their own companies or to work for competitors. Even the co-founder of DeepMind has left...for Microsoft. Morale at Google is low because their AI team can no longer work autonomously. The AI team is now forced to make productivity products instead of basic research. Their monopoly on search is now seriously threatened and their revenue stream depends on it. At best, they'll be one of a myriad of competitors. But their internet/AI monopoly is over. They have no moat. It's time for new players to take the stage and build the future.
I only see a flamboyant bombastic presentation by people with too much ironing spray starch and without any exciting product that invites us to dream. This company has definitely lost its shine.
So basically Google is advertising their eco system. I'll try not to be negative about it, but if they do apply Ai properly to all the companion services they provide will they finally hold a significant spot in the Ai universe. And I hope you haven't missed the fact that they don't have to dominate the Ai market when it comes to retail, they have other Ai companies paying them to use their resources. They sell to the companies and the companies sell to us. 😅 Let that sink in.
I love googs, AI Google not bard/Gemini is up and down. I recently got my 1st strike for hate speech on a video without words and ai created and very patriotic. Im not mad enough to get kid in trouble, it's sad though
So true it's a shame that a company of this size with this level of resources engages in leftist woke ideology. If they continue along that path ultimately it will be their downfall.
Am I the only one who hates this new UA-cam layout? So hard to read comments. What were they thinking? Maybe Gemini gave them the idea or coded it for them???
I like these videos, where an expert watches along with us and helps us understand the material. However, I’m pretty upset that segments were in here multiple times, instead of saving time consuming these videos are wasted a bunch of time we watching the same segments.
I agree with the other comments stating that Google hasn't delivered on the AI model front. However, I have no doubt that they'll figure it out. They nearly always do. The resources and talent are there. Ironically, they're focused, or it seems like they are, on beating context windows in other models when in all of their other products they're focused on simplicity and accuracy. From what I've seen and worked with, it doesn't seem like google, but I'll continue to keep track of their progress.
I don’t think they need to really innovate or be on the front lines of the model battles, they’ve realised that Opensource models are going to stay competitive and push those boundaries so they’ve just built the megastructure and infrastructure to harness the power of those models within their platform and make integrations easier, which I think is a very smart move from Google considering their already established Workspace ecosystem and integrated services
In my view, these low effort videos are harming your channel. You need to be more involved, or just don't post stuff like this IMO. You did not impart any info. Same as posting a link to the video.
Gemini advanced turned out to be horrible. However, when I’ve tried Gemini 1.5 it’s worked wonderful. I’m just waiting for it to be available outside of the testing area
These would be amazing products if the AI was reliable, and that remains to be seen. But please stop reposting videos as if you were adding something! I already saw this, and came here for your take, then just wasted my time!!! >:-(
No one: Google: I'm super for-realzez this time guys, why don't people just believe whatever we tell them, ridiculous! I pinky promise my new super AI, Fugazi powered by our Rim-job model, will be bestest ever and deff unlike anything ever seen prior. P. S. Starting in 2 days Google search will be shutting down permanently.
Too bad gemini 1.5 pro really sucks at giving you decent full code. Glad they releasing the audio though. Lmao at the "make a contribution on my first day." No, there wont be a first day, they'll just make senior devs do more while they hire less then complain when there's no seniors.
I have big questions for Google:
- Can they make a 7B model that dominates all other 1B models?
- Is it ok to be white?
I didn't realise Google was doing comedy these days
"Google finally announces huge advancements in AI. Yet again. For real this time. Promise. Dude this time is different. Trust us. Ok?"
anyone who buys their bullshit lived under a rock for some reason :D
Gotcha
Ok I'm buying gpt4 subscription 😂
Google announcements doesn't mean a crap now. I will believe it when it works for real in my hands.
Facts
I don't understand why google doesn't hire some people well versed in public speaking to do these showcases instead of putting out some monotone executive robots who move their hands in an animation loop. It was the live equivalent of when you ask AI for a concise answer and it hits you with six paragraphs of its historical origin before answering.
@Est_Since_86 Oh don't be sorry, I am however saddened over your poor reasoning skills! They are not answering questions from the public, so they don't need to be experts in the matter. They are however speaking publicly, so they do need to be expert communicators.
@@AINEET Even then, I'll never understand why people say being a poor communicator isn't a problem because "ThEy KnOw ThEiR StUfF". If it's hard to understand and follow you're just putting a meaningless problem in the way of your message.
@@AINEET no, most demos were not live (cached), i don't trust this BS
One can only hope that the speaker is AI, then it would be an incredible demo 😂
@@Luizfernando-dm2rf Yep. And let me add to that: Steve jobs wasn't a tech expert and he basically set the standard for these. Creating the product, leading the team, explaining what it does and how to use it are all entirely different skillsets. And these presentations aren't even in depth technical explanations, they are literally showcasing products everyone will be able to use lol
Me: Gemini, describe what black Nazi's did in WW2.
Gemini: on It
Try googling it. You'll be surprised.
The only thing positive is that we can expect others like Openai to have more than 1 million tokens context windows pretty shortly. That's what I heard.
I'm not listening to all this corporate garbage. Someone give me the rundown of anything substantial which was said.
The TLDR is that google is all hype and no remotely competitive AI product. Par for course of course.
You'll be getting more than 1 million tokens with the next openai or anthropic or something like that.
@@thomassynths Thank you. :)
I will translate: Woke, woke, bla, bla, woke, woke... 😂😂😂
Nothing inciteful. All hype, again. And models available for public use are still garbage.
Ugh. Google's just all hype now. Remember their so called ChatGPT killer called "Bard"?
Gemini is pretty good, though. I use ChatGPT and Gemeni for copy writing and Gemeni often comes out on top
Bard was the worst piece of shit ever.
RIP Bard. They "discontinued" it.
Where do they get these names from? Bard, Grok, Llama etc. They all suck.
@sentry8992 there is this realm where demon ai are wrangled into being hood little ai. They come named already. Hard to get them out of that layer of hell though, this is why Christians fear ai.
I am rather scared of google advancing AI. They have the most biased LLMs out there
Gemini 1.5 pro. Now 1.5 times as racist as our previous model!
Was trying to summarize comments on a pro life blog talking about abolishing abortion, I kept getting an error- too many tokens generated.
An some article summaries i got "is a complex one with no easy answers" skipping the details in the article.
The Censorship is wild 🎉😜
As humans start to solve racism among themselves, there's a demand for more racists to fight, that's were Gemini comes in
@@Luizfernando-dm2rf It’s possible. That’s the only sane explanation I’ve heard for the far-left ideological bias and woke racist world view which is baked into every AI to some degree.
F@#& Google
Not going to lie - I laughed at loud at that one. 🤣
Gemini 1.5! Now 50% more woke and racist! Still completely incompetent and unusable!
Google has been incredibly stupid in their training, but they have way more compute than anyone else, so I’m not counting them out. (But I’m rooting hard for everyone else.)
Why are you "rooting hard" for others? To me that seems incredibly stupid, if not fanboyish 🤔. It's multiple comments like this that lead me to believe people are honestly just sheep. So with that in mind let me tell you exactly what's happening here:
Baaah. Baaaaaah. BAAAAH! ...baah. 🐑 😏
@@Cross-CutFilmsBecause Gemini sucks? Why shouldn’t we root for the competition? The only sheep here is you. baaaaah.
Ram too ♈
@@Cross-CutFilms you didn’t really make an argument about why you think rooting for Google’s competitors is a bad idea. But you made a whole string of animal noises and used several emoji, so well played sir?
@@Cross-CutFilms Google has shown itself over and over again to bias results away from the truth in favor of their preferred narratives. (Look up Robert Epstein’s research for very well-done research showing this clearly in Google’s search results. Look at Gemini”s recent laughable refusal to make images of white people, showing everything from Vikings to Nazi youth as people of color.)
If you think we should trust Google and get behind whatever they’re doing, I’m not the sheep here.
I guess Google is now an Indian company.
They were never American- globalists.
Yes, and let us give google all the extra data for free, while paying em for taking it. Yes, great stuff, but data?? What the quack!
Mr Berman...why is this on your channel without any commentary or analysis? I know these are trying times economically....
Presumably an AI edited and uploaded this video without human involvement.
He edited it. It’s super useful.
I think Matt wanted to share this but not comment on it or take any ownership of it. I can't imagine why!
It’s better to bite-your-lip occasionally.
It all sounds great and then I think about crap Google products like docs, sheets and Gmail. Makes me not totally bite into the hype.
I have very little faith in Google's ability to deliver AI innovation, regardless of their claims. Their actions, not their words, are the only thing that will convince me to be stunned and shocked in this area.
Yep, they talk a good deal but keep falling short, which is surprising considering Google were ahead of the race for many years but Google never took advantage of that by not releasing any to market, they talked a good deal in the lab, but very little was seen on the market.
But honestly, I'm far more interested in open source A.I. and especially the ones you can run locally, I think that's the future of A.I. and over the last year, I went from being very interested in the closed A.I. solutions to hardly caring about them at all that I don't watch videos or read tech sites on them any more, I've become far more interested in open source A.I. because there's a lot more potential and because no one controls it and it's become more the case with how closed models are behind a monthly paywall, which as good as some of their solutions might be, I've got little to no incentive in wanting to use them with how much they charge, especially with how quickly alternative open solutions keep developing.
@@paul1979uk2000 Open source models have been growing on me too and I am ready to make the leap. Are you using cloud computing to run your models or did you invest in a home workstation?
Unless you're a developer or an institutional investor, your opinion accounts for little
@@IbukunFisher True!
To put in the right context, Google has to take the safe approach to AI, at risk of jeopardizing their image when things go wrong. Many other players had no market share and are trying to disrupt the market with experimental features and get little global news coverage when things go wrong, including Microsoft initially. To say that Google is behind in AI would be saying that Tesla wasn't ahead of the competition in electric vehicles just because some small startup had one novel feature that was better, when it worked the 30% of the time.
I haven't used Gemini, but I added the extension to vscode to try it out. I asked it to analyse my existing code, and it analysed and provided suggestions. I then looked at what they did in the demo by pressing the + button.... and it doesn't work. It doesn't update the specific line of the code it just randomly places its suggested code to wherever the cursor is in the code window. This is different behaviour to what they showed in the "demo". I'm sure gemini is great, and code agent isn't properly released yet but that extension doesn't event give my the line number....I
m going to take a wild guess and say that the demo they showed is BS.
the "DEMO" is not live, it was cached, they mentioned it too.. saaaad.
@@hqcart1 She clicks the + button to add it, that's not how the plugin works. You have to first find it, then highlight or remove it then press the plus button. That part is not "cached" that shows a completely different experience. Gemini doesn't even provide the line number of where its referring to.
Holy crap she showed a decision on a system replacement without reading the document...wtf...
google appears to support the future total absence of any human critical thinking. And LLMs are not 100% correct
At least Google are experts in wideo.
Matt, we need you to go through this and tell us what you think in the next video
It’s would be disingenuous to be excited by Google.
Gemini was force fed into Google assistant and it absolutely sucks. Google doesn't respect its customers in the slightest.
Competition is good. But The switching cost/effort for users is too high, so OpenAI/Microsoft won't lose its position anytime soon.
Depends on your architecture. For me, switching would take... 1 day? But switching to Google is not an option. They have zero trust.
A message from our Evil-Corp overlords 😕
Decades ago, computers took up an entire room. We are at the start of AI transformation. Things will only get better as the tech advances.
Matthew thanks for the supercut.
As you know, I personally feel that any company not looking into analog AI or hybrid analog/digital AI is running down a dead end road.
Analog is better than digital for differential calculations and beats digital by orders of magnitude on the power consumption per compute curve. Then you need to consider the potential of AIs based upon continuous multilevel logic.
Released on April 01 :))
Am I the only idiot who initially interpreted "dropping" as "abandoning" in the title and got really confused?
hmm...i hope that isn't confusing for more people...
It is still a fairly new usage of the word. While it is more and more common these days, time will tell whether it sticks in the language or becomes dated slang.
I’m just always glad to see a UA-cam video without the term “Deep Dive” or “on another level”, especially when it is in every title on every video on a channel. “Deep dive” is definitely a term whose time has come and needs to go.
the insurance claim automation sounds like the premise to "Brazil"
They sound likr corporate robots. I'd hate to work with those people
"Gemini ensure that code is aligned with security and compliance" .... this is a somewhat ironic statement, as most software companies security policies are VERY clear about not having their entire source code sent to someone like googles server to be scanned and processed.
That’s an awful lot of hand waving going on in this video. All I see is Google back where Microsoft was in 2007. No longer competing. It’s wild to see what OpenAI has done to these tech companies.
I have trust issues after the Gemini rollout
Thanks Uncle Sindar for the app ideas 😉
The choice of model sounds interesting. And maybe that’s what Apple is doing. They are running their own model on Google‘s hardware.
Original AI Google great, bard has been inhibited sadly, I have been using aicroq because of Google bias
Deployed Worldwide Through My Deep Learning AI Research Library … Thank You
Couldn't help but notice Visual Studio code icon was on the choice of development environment slide but it didn't have a name next to it like all the others did. Hehe
When their marketing team uses AI to analyze the sentiments here, it will be another awesome launch
thank you Matthew for the impressive selection of contents that you are sharing on AI and its application. What is really interesting in this video is the total absence of any critical thinking from the human beings (?, one has to wonder...) about using LLMs tech to make a decision in their behalf on really important personal matter like health insurance. I am saying that the message 'Gemini can filter and compare CORRECTLY the relevant information and thus fabricate a specific reality for the human being where humans would want to agree" is quite scarying and totally misleading given that LLMs are not right every time. And just a 10% error is really impactful on these real life matters.
With more data accuracy will improve
As Google search continues to erodes, as it has for years, that calls into question all his blathering. If it doesn’t work in house (and maybe a driver of the decline itself), why would another entity use their tools (except for basic storage and runtime). What a joke. Zero credibility. You can run experiments on Google search yourself, and see that in many cases, whether for technical subjects, or even their bread and butter consumer search, for all but the most mainstream, it often just doesn’t work. It’s like going back 20 years, where you need to find proprietary sources and do a lot of manual work.
I feel like I'm getting flashbacks to prior tech companies before their decline.
I'm thinking of IBM, Atari, Commodore or even MS.
It's this mishmash of business jargon and smattering of applications thrown at the wall, in the hopes of landing the next big thing without the singular, coherent vision.
The tone of the pitch sounds subtly desperate. They already control so much of everything else, there would be no shame in them not trying to own AI as well, and just supporting it with all their other tech. But they can't help it.
It's so sad to know we're in a breakthrough period with advancements that will improve human quality of life and the richest companies have appointed out of touch executives with no coding knowledge to step down from their thrones and tell the public how amazing they are. I guess all that's important anymore is who's getting more money.
I'd never use google AI, they will steal all your info... PASS!
720p lol what year is this
I don't think Google would success with this CEO. I would personally never invest on Google till they change the actual CEO.
I am very lucky to see high-level sharing and benefit a lot.
Does he not seem like the guy that would also be on a big screen telling you why futuristic prison camps are good for you and the obvious path forward. 0:16
Great announcement, thanks for sharing!
I've been working with several of these offerings over the past few months, developing some great use cases. I was thinking of perhaps showcasing some of them on my channel soon.
They are incredible at knowing the right things to say. Until they show me something meaningful I’m going to live under the assumption that google is a follower in this space and nothing more.
water looks like a 2000's game..
Matt, I am from Missouri! After the quacking ducky episode, I'll believe it when I see it!
That TTS demo audio was so bad that he had to stop it.
Where is Apple?.....ssssshhhhhhh....we wait
They are not going to need to water mark this stuff, it is so bad
If I had watched 2:05 to 8:04 just three years ago I'd be scratching my head throughout the entire thing. :D
I'm still scratching my head for some of these terms.
7:10 - "For your most stringent regulatory....you have control over your data..."
Yeah that's the way it should be for everyone, not just the most stringent.
9:34 - Can Gemini set reminders on my Android phone yet?
21:00 - Did anyone ask for digital watermarking? Can't wait for the digital watermarking removal demo.
Dang, talk about lack of applause lines. This audience is not pumped.
They are all scrolling on their phones (being paid to endure it)
24:05 the picture is bad, there is no rain tarp on the tent, it isn't staked down, the mountains are blurry and off, there are no hands, etc.
Lol I heard that marketing BS many times, I’m not listening to it. Google cope better. Big L
It's surprising to see so many negative comments. I have no association with Google particularly, so have an impartial view. Fairly speaking, Google made the most fundamental contributions to the current AI boom, including theories, tools, and people.
So eventually it will belong to Page & Brin?
Aren't those two flying fighter jets to their privave aircraft carriers moored off private island on a run on sentence?
Google has failed to deliver on any of their promised products for a couple of years now (if not more). I'll be waiting for a free trial to see for myself if Gemini 1.5 can finally reach GPT-4 capabilities.
This "supercut" isn't super enough. Over half an hour of this overhyped corporate jibber-jabber and underwhelming contrived demonstrations is too much to bear. Oh, and announcements are made, not "dropped".
I feel a deep hatred for that company and for that guy's accent
How about being able to export gpt conversations with embedded images into second brain-,type apps?
I'm not sure what HUGE means in the context of this supercut. But maybe you put it at the end.
What is the best LLM to run locally for screening scientific articles? I mean it should have great reasoning and understanding of english (scientific) language.
The competition in the field of AI is growing rapidly. Talent is leaving Google now to start their own companies or to work for competitors. Even the co-founder of DeepMind has left...for Microsoft. Morale at Google is low because their AI team can no longer work autonomously. The AI team is now forced to make productivity products instead of basic research. Their monopoly on search is now seriously threatened and their revenue stream depends on it. At best, they'll be one of a myriad of competitors. But their internet/AI monopoly is over. They have no moat. It's time for new players to take the stage and build the future.
There is so much AI hype, and they cannot correct the echo in the audio. . .
Google-run Privacy Preserving AI Models. Yes. They advertised this BS too.
At Google.
GTP5 needs to come out soon or they'll have egg on their face
I only see a flamboyant bombastic presentation by people with too much ironing spray starch and without any exciting product that invites us to dream. This company has definitely lost its shine.
So basically Google is advertising their eco system.
I'll try not to be negative about it, but if they do apply Ai properly to all the companion services they provide will they finally hold a significant spot in the Ai universe.
And I hope you haven't missed the fact that they don't have to dominate the Ai market when it comes to retail, they have other Ai companies paying them to use their resources. They sell to the companies and the companies sell to us. 😅 Let that sink in.
How good are Google's Coral AI TPU USB boxes? Do they perform well?
24:39 if he can speed up time, the whole thing is pre-rendered.
Damn, I thought this is Google's channel
I love googs, AI Google not bard/Gemini is up and down. I recently got my 1st strike for hate speech on a video without words and ai created and very patriotic. Im not mad enough to get kid in trouble, it's sad though
Dude , I meant this channel broadcasting this entire content, it's not made by him but he's posting on his channel and getting views.
I agree, but no words were said, Trump was totally Lisa franked, and song was attached in studio
Nobody is watching that
The question is: in Google’s estimation, did they improve their previous woke racist AI by making it more woke and racist or less?
So true it's a shame that a company of this size with this level of resources engages in leftist woke ideology. If they continue along that path ultimately it will be their downfall.
No. Now it supports 1M tokens of wokeness 🎉
I personally don't care about that. I just hope it codes well.
@@coom07 that’s a lot of black Nazis and Vikings😂
Designed to better hide the bias and racism
This is just a 'supercut', no commentary? If I wanted to watch the video, I would just go to their channel.
I bet all what we've seen was not generated in real time. Google can't afford another flop.
90% using Google Cloud… don’t know a single dev team using Google’s AI. Not one! Total horse sh.t
I hate that you just upload these videos but don't give any thoughts or extra insights
It's a move borrowed straight from WesRoth's playbook.
Am I the only one who hates this new UA-cam layout? So hard to read comments.
What were they thinking? Maybe Gemini gave them the idea or coded it for them???
I wish you'd had some commentary here and there in the video, Matt.
I like these videos, where an expert watches along with us and helps us understand the material. However, I’m pretty upset that segments were in here multiple times, instead of saving time consuming these videos are wasted a bunch of time we watching the same segments.
I agree with the other comments stating that Google hasn't delivered on the AI model front. However, I have no doubt that they'll figure it out. They nearly always do. The resources and talent are there. Ironically, they're focused, or it seems like they are, on beating context windows in other models when in all of their other products they're focused on simplicity and accuracy. From what I've seen and worked with, it doesn't seem like google, but I'll continue to keep track of their progress.
I don’t think they need to really innovate or be on the front lines of the model battles, they’ve realised that Opensource models are going to stay competitive and push those boundaries so they’ve just built the megastructure and infrastructure to harness the power of those models within their platform and make integrations easier, which I think is a very smart move from Google considering their already established Workspace ecosystem and integrated services
They have Mustafa now so they will do well
Yawn. Hardware means nothing when they are purposefully gimping their LLMs to protect feelings.
In my view, these low effort videos are harming your channel. You need to be more involved, or just don't post stuff like this IMO. You did not impart any info. Same as posting a link to the video.
How can anyone endorse this guff… and how can anyone be against it…
@@alekjwrgnwekfgnnot sure what you mean
As if companies are going to send a photographer around the world to shoot on location if you can get this out of the box :)
Gemini advanced turned out to be horrible. However, when I’ve tried Gemini 1.5 it’s worked wonderful. I’m just waiting for it to be available outside of the testing area
meanwhile openai buy more 4090s lol
what good is any of this tech if it's forced propaganda
AI doesn't like LIES so please stop trying to make it "Woke".
I am too drunk to understand this.
These would be amazing products if the AI was reliable, and that remains to be seen.
But please stop reposting videos as if you were adding something! I already saw this, and came here for your take, then just wasted my time!!! >:-(
No one:
Google: I'm super for-realzez this time guys, why don't people just believe whatever we tell them, ridiculous!
I pinky promise my new super AI, Fugazi powered by our Rim-job model, will be bestest ever and deff unlike anything ever seen prior.
P. S. Starting in 2 days Google search will be shutting down permanently.
What the fuck is an AI unicorn?
everyone is talking so fast, did they speed up the play
Too bad gemini 1.5 pro really sucks at giving you decent full code. Glad they releasing the audio though.
Lmao at the "make a contribution on my first day." No, there wont be a first day, they'll just make senior devs do more while they hire less then complain when there's no seniors.