I mean it's not like Google was not at the origin of most of the technical advancement that allowed the creation of LLMs 😂 They just got surprised by openai releasing a product and a tech that was not really ready for prime time. Frankly Google has been held to a double standard until now regarding hallucinations of their LLMs as opposed to openai's. Where you had Google being smacked down for it while openai would get a free pass while doing exactly the same kind of mistakes. And this is probably what has been slowing down Google so much on AI topics and advancement specially when it's about releasing the thing to the public
They have the training data (the world's search engine and advertising data pretty much), the compute and the money to do things with that data, so it is not that surprising. If anything the surprising part is that it took them this long.
Not surprising, they are the ones that designed how AI works, they just kept all the fun stuff as experiments until other players swept the rug off their feet
42:29 The diversity of things that you can do with Gemini it'll be absolutely life-changing to everyone. I must review this episode over and over again because it is full of practical usages of Ai utilities. Gemini and others will take over the world.
@26:00 It actually got the perimeter math problem wrong. It missed the vertical side, which would have been 6, making the computation: 12+9+3+(6)+9+3 = 42 The AI is still advancing, so be careful with these early stages ~
That's an understatement! I'll find a good tool I like and then buy a monthly subscription and before the subscription ends there'll be another better tool in the same class that's available.
Would you go over example question and workflow. I have been wanting to use LLMs for studying however it feels like pseudo-intelligence and like I skip steps when trying to learn with AI. Best way to explain is using wristbands/knee bands to lift heavier weights. What happens when you have to lift that same weight without the support? Keep in mind I only have this trouble in proof based mathematics. Simple calculation math is not an issue.
This is probably the one video I never wanted to finish watching. Thanks so much for your effort-it truly means a lot. Hearing all this news together was incredible. Really, great work. Good job man 👏
Most people still don’t realise how much AI is going to change our almost every aspect of our everyday lives, even in social contexts. For better and worse.
The vision capabilities would definitely be useful for those who are blind like me. Just amazing that AI has come so far already. The fact that this is free is amazing.
Nothing is truly free. They use your activity in their services to make profiles of you and sell them. They use the videos you post to train AI. They own the writings you post on social media and sell them if they want. And on top of that they make more money showing you ads and selling you premium features. For every dollar of " free" stuff they give you they make ten times that. But yeah. I also love to be able to use some AI for free, despite all the limits I hit daily.
@@ronilevarez901 well at least it doesnt cost me nothing. If they make money making people's lives better then they can have all the money they want. If they're not worsening our lives, then let them have their ego games and political plays and whatever. We need only to worry when they worsen the world more than they better it. And this is a great leap towards the good side of the scale
@@bcrox why are you on every comment being so pessimistic 😂 bro really watched 43 minutes worth of a great product to bad mouth it. I have a secret for you, everything you say is already being listened to from your phone. Your location is always being tracked. Everything you type is saved and used. So havhbg something that actually helps you with the information they take FOR FREE anyways, is good already. Go back to a type writer if you're so scared of the future and advancement 😂😂
Google jumped ahead of the pack this last week. Its pulling farther away from everyone else in AI. Insane AI features all across its apps, products, devices. And its only just begun.
Tbh gemini 2 is way better as a teacher than any other teacher who isn't like a guide and just yapp bout what's in the textbook and leaves even if it is like $100-150/year it's worth
It's basically life giving us a chance that we've never had unlike those who've had great mentors/teachers throughout their childhood/lifetime. Now we all get access to it and can essentially become similarly minded as those who've had such great people in their lives
Google's AI products are garbage. You're really getting cheated by sticking with this. You should absolutely do your own comparisons with OpenAI and Anthropic's models, and you'll see.
@@zenphoenix yes i used them but they have less context length, so in this you can summarize all formulas of your notes, and yes it is inferior in quality but its free and unlimited
4:40 This is insane. It even responds like a human now. "Great......Ok, I'll wait". I mean.....I'm speechless. If anyone had played this conversation without telling me there was an AI involved, I'd be thinking it was just two dudes talking to each other.
24:20 Gemini has taken over everything, we now can use Gemini to assist us with everything. Helping with software tutoring, with coding cases and giving proper information on medicine usages is just next level capabilities. 25:35 The translation speed is super fast, it's on the comic character flash speed.
As a MD it was better than me for seeing the dental cavity. My attention was focused on the gingival retraction (which the AI did not indicated BTW). Impressive and thank you for presenting this fascinating video and information.
"What will you build. Push the limits of Gemini to see what you can do." Yeah it's obvious what this is. With that being said, I stand behind it 100%. Using AI to improve AI in a realistic way is when the paradigm shifts again, and hopefully, this is a major one. I'll use it for my next project, which is simple enough; very specific song lyrics that I can feed into Suno and make something 99.9% of people would have no idea had any AI involvement in its creation. It's not creative on my part, really, but if it can write lyrics better than me for the style I use, I would be absolutely floored.
One complete failure test I found: I tested its musical ear and it does not have one lol. It thought an octave was a perfect fifth, it called a tritone a major 2nd, it thought Stairway to Heaven was Twinkle Twinkle Little Star, and repeatedly misidentified chords - seemingly no better than chance choosing between minor, major, diminished or augmented. Then I just asked "is this a high note or a low note" and again - no better than chance. I asked if it could identify my age or sex by my voice and it couldn't do that. It almost seems as though the language processing and audio understanding are completely separate.
Update: I did the same with ChatGPT and ran into the same issue, although it was much better at explaining why it couldn't accomplish the tasks. "while I can chat naturally with you here, I don’t have the capability to directly process live audio like your guitar through this setup... The issue likely comes down to how my system is designed. While I can process your voice in real time for conversational purposes, I don’t have the ability to interpret audio streams like musical instruments. My focus is on understanding spoken language rather than analyzing the full spectrum of audio inputs. It’s like having a radar for speech but no sonar for music-different tools for different tasks! If analyzing live guitar playing becomes possible in the future, you’ll definitely be hearing more from me on that."
It's because of your microphone which is noise removed and is what generally normal AI does because of interruptions of its Voice by its own @@LydianMelody
@@LydianMelody we're definitely gonna overcome those problems relatively soon, hallucinations are still a thing, and text is much more explored and optimized than audio so lets just give it time
so this is what i get when i tried using this: me: "can you see my screen?" AI: "I can't see your screen as I am a language model and don't have the ability to see or interact with your screen. Is there anything else I can help you with?"
I'm jealous of younger people who can use tools like this to learn like never before, unlike me back in my day only having textbooks, youtube didn't even exist when i was in school, and if i got stuck on something and textbook wasn't helping me, i was literally stuck.
Same. Back then I solved it the simplest way: I didn't learn, lol. I eventually quit school and learned more then than with teachers. Knowledge is pretty useless without the certification from School but still, it's another proof of how outdated the scholarized system is.
I love how the AI gets the perimeter question wrong, it's D)42, they missed the 6 from the short side. AI will always confidentially declare the answer, even if it's wrong. I love AI, but this is the part that is scary to me.
Thanks for this. I hope it’s true. Can you demonstrate using it to setup and build a Shopify store… fast and effective to ramp up SEO and monetization? This might be the AI democratization we hear so much about “Prompt my computer”.
It doesn't work for me, when I share my screen, it keeps saying it can't see the video and it is a language model without video sharing capabilities. Even though the video is already in the chat log screen, it still can't see it for some odd reason.
Finally something useful I always hate the idea of artificial intelligence doing all our work but this time it has become a tool to improve our skills and solve the problem of difficulty in understanding I can see myself studying something new and difficult Imagine yourself holding a book and wearing AR glasses and asking the AI how can I start reading this book and make a plan
I hope that the api stays free for developers forever ! This is amazing im able to make so many good stuff for my personal work and stuff. This is insanely smarter than gemini 1.0 pro api
Thank you for going over this. I look forward to testing the upgraded voice options when they are released next year. The real time conversation option with Gemini 2.0 Flash Experimental seems to error out after chatting for a few minutes, at least it did for me with a dozen attempts.
6:15 This is like having a Professor right by your side guiding you through a project or lesson, and who you can question whenever you please to expound on difficult topics. This is clearly a game-changer. 😎💯💪🏾👍🏾 12:02 Is this for real or are they faking it? Because if Gemini can interpret medical tests and make legitimate medical diagnosis then that's very impressive. I'm lost for words.
Google has obviously been sandbagging until the other players had played all of their best cards. Now all eyes will be on them as the gain all of the market share in AI products just before they become flush with cash after being forced to sell chromium and android before the end of the summer.
Thanks for the review... Any concerns on security from passwords and confidential dsta when sharing the screen? What about Keeping projects/conversations memory?
I just want to prevent companies from training on the data I input or use it in any way and tbh no matter Google says, I wouldn't trust that they won't use my data. Their reputation screams.
well i guess thats a concern, people that manipulate that kind of information should wait to the point when these kind of models are available for anyone to download and run locally and modify them all they want, until then, the information you share reaches someone's servers
I wish Google Gemini 2.0 could be integrated with UA-cam soon. I showed Gemini live a song I listened to on my TV and it correctly identified it. I asked it to play similar songs on youtube. It could not, because Gemini 2.0 is not integrated with UA-cam yet. That would be powerful to put Gemini and UA-cam together. Just imagine all the possible use cases.
Is there an AI voice speaking on this channel? It sounds so strange when every other word is emphasized and there is a short pause between all the words.
im pretty sure they are working on something like this. my dream would be this ai implemented in new tvs, watching movie or tv show pausing for second to live ask ai what is that actor how old is he, or where was that scene filmed, what brand car is that etc etc, i spent tons of time doing that on internet, live chat would be so cool
@@theAIsearch i used it to react to me playing skyrim in realtime and every 5 minutes or so it stopped and said "something went wrong" i wonder if its just because they just released it and its experimental, but also they give you less control when you hit the show option in Stream Realtime option. Also in my particular case the AI was too explainy with everything it saw on screen, so it would again be amazing to have some more control of it, i hope they add that
google really has unlimited potential at this point, they are making everything from chat bot, to live chat, to video chat, various ai video and picture generators, not to mention they have their own smartphones, earbuds, smartwatches, imagine this all implemented and optimisedin google pixel ecosystem? they have google powered search that is still top even tho it is infested wit ads and ai stuff but still it THE search engine, add youtube , yt music etc... google might take over everything soon, asking your smartwatch or earbuds to help you do anything in your life, from learning ,teaching , working to simple things at home... this is real main competitor to chat gpt imo, unlimited potential if they get it right. even now its super impressive.
with only this, *not 'only' actually, but you know for 'one' new AI model, Google literally crushed OpenAI (with their not so new 12 days live program haha)
Have you seen if the use of a webcam or the view of your desktop functionality is available programmatically through an api? I would love to incorporate this functionality in my application.
I admit that I was a skeptic, but it does increasingly seem like there is no wall. All of these companies keep moving forward. Sure, there is a leveling off in classical LLM transformer architecture with scaling, but overall, AI continues to race ahead with clever new methods. We do seem to be approaching something like the sci-fi AIs at this point. Unclear still if we can hit an AGI that truly has novel ideas and can advance science forward on it's own, but as a tool to efficiently access all the knowledge of humanity, it seems like we are there already.
I'm totally blind and will always trust my ears to know when a light is green. This has all kinds of potential navigation uses I'm eager to test but you picked about the worst possible one. Even if it's seeing everything clearly, even if you have a good connection there will be just a little lag and, ya know, talking to it while I need to be listening to my traffic might just get me killed. Great idea, awful example.
Two months ago i been to see doctor because i droped something heavy in the hand, i only go to doctor if i realy feel that i have to. so doctor at the hospital did x ray of my thumb and his diagnosis was that there is nothing wrong with bone no fractures nothing and everything is ok. today after two months i still know there is something wrong with my thumb cause i feel pain inside. so i uploaded image to gemini and there we go it said ,,it apears to show a fracture of the distal phalanx of the thumb. specifically there seems to be transverse fracture line visible across the midshaft of distal phalanx. Im very angry that doctor did not see that but gemini saw it straight away. Thats crazy
I was teaching Google voice AI how to laugh and it broke down as a result. But when I made a new voice chat, and asked it to laugh for me, it laughed properly whereas before, it didn't work. Freaky...
It really is, revolution in learning, no? Schools are perhaps no longer needed. It used to be " yeah schools won't even be needed once it's advanced enough" WELL NOW WHAT? This is so crazy
it will be awesome when it actually usable, no shade, but honestly can't use it more than a few minutes before it quits and I have to start all over, unable to train either afaik.
This is pure insanity to be honest. Google went from being a mid-player to one of the most impressive.
yeah they've released some insane stuff recently
I mean it's not like Google was not at the origin of most of the technical advancement that allowed the creation of LLMs 😂
They just got surprised by openai releasing a product and a tech that was not really ready for prime time. Frankly Google has been held to a double standard until now regarding hallucinations of their LLMs as opposed to openai's. Where you had Google being smacked down for it while openai would get a free pass while doing exactly the same kind of mistakes. And this is probably what has been slowing down Google so much on AI topics and advancement specially when it's about releasing the thing to the public
Fr
They have the training data (the world's search engine and advertising data pretty much), the compute and the money to do things with that data, so it is not that surprising. If anything the surprising part is that it took them this long.
Not surprising, they are the ones that designed how AI works, they just kept all the fun stuff as experiments until other players swept the rug off their feet
The "AI" we imagined years before! This is so human like and like the ones shown in sci-fi movies... mind blown
i know right, basically cortana, imagine when this thing incorporate in videogames
@@upsidedownhorror😳😳😳
@@upsidedownhorror imagine when this thing incorporated in everything collecting everything you See, Do or Say 👍 super cool
42:29 The diversity of things that you can do with Gemini it'll be absolutely life-changing to everyone. I must review this episode over and over again because it is full of practical usages of Ai utilities. Gemini and others will take over the world.
@26:00 It actually got the perimeter math problem wrong. It missed the vertical side, which would have been 6, making the computation: 12+9+3+(6)+9+3 = 42
The AI is still advancing, so be careful with these early stages ~
One day the AI will know that the answer is 42 🙂
Went into the comments to see if anybody else also caught this
Crazy fast improvements in AI. It seems like every week brings what recently would have been a decade worth of advancements.
That's an understatement! I'll find a good tool I like and then buy a monthly subscription and before the subscription ends there'll be another better tool in the same class that's available.
yeah, this month has been crazy
I wouldn’t be surprised if AI would be able to make GTA7 on its own around 2030.
@@gameboardgames😂😂😂so true. I have check mine like every other week
@@theAIsearchDecember usually is a big month for AI releases as the past couple years have shown.
I used gemini 2 to learn maths, i loved how i could either ask for next step or just hints and it'll give me hints
Same as for me. We are truly living in the most beneficial time yet. God bless! ❤
thanks for sharing!
kaise access Kate bhai
Would you go over example question and workflow. I have been wanting to use LLMs for studying however it feels like pseudo-intelligence and like I skip steps when trying to learn with AI. Best way to explain is using wristbands/knee bands to lift heavier weights. What happens when you have to lift that same weight without the support? Keep in mind I only have this trouble in proof based mathematics. Simple calculation math is not an issue.
I used the screenshare feature to help me fix a problem on my computer... and it did! 😊 Awesome Tech
nice! thx for sharing
what problema was?
effici AI fixes this. Google's Gemini 2.0 AI released.
This is probably the one video I never wanted to finish watching. Thanks so much for your effort-it truly means a lot. Hearing all this news together was incredible. Really, great work. Good job man 👏
Wow, thank you!
I got heavy into Gemini just before 2.0. It is an absolute beast paired with all of Google's tools.
Most people still don’t realise how much AI is going to change our almost every aspect of our everyday lives, even in social contexts. For better and worse.
The vision capabilities would definitely be useful for those who are blind like me. Just amazing that AI has come so far already. The fact that this is free is amazing.
Nothing is truly free.
They use your activity in their services to make profiles of you and sell them.
They use the videos you post to train AI.
They own the writings you post on social media and sell them if they want.
And on top of that they make more money showing you ads and selling you premium features.
For every dollar of " free" stuff they give you they make ten times that.
But yeah. I also love to be able to use some AI for free, despite all the limits I hit daily.
Schwab Klaus said whoever controls A.I tech will control the world.
He also wants global population to 👎
@@ronilevarez901 well of course. If something is free, the data that you give them is the cost. I’m well aware of that.
@@ronilevarez901 well at least it doesnt cost me nothing. If they make money making people's lives better then they can have all the money they want. If they're not worsening our lives, then let them have their ego games and political plays and whatever. We need only to worry when they worsen the world more than they better it. And this is a great leap towards the good side of the scale
I want this in my glasses and airpod imagine how this going to change my life
Yeah, imagine it collecting everything you see, do or say forever, your life would be changed. 🤔
@@bcrox why are you on every comment being so pessimistic 😂 bro really watched 43 minutes worth of a great product to bad mouth it. I have a secret for you, everything you say is already being listened to from your phone. Your location is always being tracked. Everything you type is saved and used. So havhbg something that actually helps you with the information they take FOR FREE anyways, is good already. Go back to a type writer if you're so scared of the future and advancement 😂😂
Google jumped ahead of the pack this last week.
Its pulling farther away from everyone else in AI.
Insane AI features all across its apps, products, devices.
And its only just begun.
Tbh gemini 2 is way better as a teacher than any other teacher who isn't like a guide and just yapp bout what's in the textbook and leaves even if it is like $100-150/year it's worth
very true
And still cheaper than a human teacher...
It's basically life giving us a chance that we've never had unlike those who've had great mentors/teachers throughout their childhood/lifetime. Now we all get access to it and can essentially become similarly minded as those who've had such great people in their lives
Real (im at uni)
i am using Gemini all day now , its incredible, and i am in high school and it helps me a lot 🙃🙂
Thanks google for doing our homework
Google's AI products are garbage. You're really getting cheated by sticking with this. You should absolutely do your own comparisons with OpenAI and Anthropic's models, and you'll see.
@@zenphoenix yes i used them but they have less context length, so in this you can summarize all formulas of your notes, and yes it is inferior in quality but its free and unlimited
4:40 This is insane. It even responds like a human now. "Great......Ok, I'll wait". I mean.....I'm speechless. If anyone had played this conversation without telling me there was an AI involved, I'd be thinking it was just two dudes talking to each other.
sometimes it answers even faster than a human
rip to all the students graduating soon who never got to use such an OP thing
I'm actually pissed, my teachers didn't know jack s***t, I could've used Ai to teach me all the things I needed to learn
Precisely, if I had these when I was a student by graduation Time I was going to surpass them @@adolphgracius9996
Lmao.
they can use it. it's available. ?????
Same, luckily i have AI now. Started Uni this year. Such a blessing. Everyone is abusing everything out of AI
24:20 Gemini has taken over everything, we now can use Gemini to assist us with everything. Helping with software tutoring, with coding cases and giving proper information on medicine usages is just next level capabilities. 25:35 The translation speed is super fast, it's on the comic character flash speed.
As a MD it was better than me for seeing the dental cavity. My attention was focused on the gingival retraction (which the AI did not indicated BTW). Impressive and thank you for presenting this fascinating video and information.
thanks for sharing!
Really appreciate the video. This channel is my favorite AI news/reviews etc. Even with some humor sometimes thanking you Sir!
Yes I use this everyday since it was released a few days ago. It’s extremely useful I use the visual capabilities all the time.
Perimeter length had 2 unnumbered lengths, and all this assumes the angles are all 90deg. Answer is not C- 36, it's D - 42
Google is killing it lately...
Real time translate from game in japanese language is huge for me, i can now play jp game without waiting global release.
thanks for sharing
Yep
Really impressed with the amount of work and range of examples you included here, will be pointing people towards it. Thank you!
25:59 Incorrect. One of the vertical sides is missing too. It's 6 units long, so the correct answer is 42.
yes
Haha the jailbreak part with the prediction was funny 😂 great content this new model is crazy good
Thanks!
26:00 it actually gives the wrong answer >.< The answer is 42, both you and the AI missed the missing vertical side.
"What will you build. Push the limits of Gemini to see what you can do."
Yeah it's obvious what this is. With that being said, I stand behind it 100%. Using AI to improve AI in a realistic way is when the paradigm shifts again, and hopefully, this is a major one. I'll use it for my next project, which is simple enough; very specific song lyrics that I can feed into Suno and make something 99.9% of people would have no idea had any AI involvement in its creation. It's not creative on my part, really, but if it can write lyrics better than me for the style I use, I would be absolutely floored.
hey, if some people like it and you make some money with it, its not really wrong, every person gets by how they can
One complete failure test I found: I tested its musical ear and it does not have one lol. It thought an octave was a perfect fifth, it called a tritone a major 2nd, it thought Stairway to Heaven was Twinkle Twinkle Little Star, and repeatedly misidentified chords - seemingly no better than chance choosing between minor, major, diminished or augmented. Then I just asked "is this a high note or a low note" and again - no better than chance. I asked if it could identify my age or sex by my voice and it couldn't do that. It almost seems as though the language processing and audio understanding are completely separate.
Update: I did the same with ChatGPT and ran into the same issue, although it was much better at explaining why it couldn't accomplish the tasks. "while I can chat naturally with you here, I don’t have the capability to directly process live audio like your guitar through this setup... The issue likely comes down to how my system is designed. While I can process your voice in real time for conversational purposes, I don’t have the ability to interpret audio streams like musical instruments. My focus is on understanding spoken language rather than analyzing the full spectrum of audio inputs.
It’s like having a radar for speech but no sonar for music-different tools for different tasks! If analyzing live guitar playing becomes possible in the future, you’ll definitely be hearing more from me on that."
It's because of your microphone which is noise removed and is what generally normal AI does because of interruptions of its Voice by its own @@LydianMelody
@@LydianMelody we're definitely gonna overcome those problems relatively soon, hallucinations are still a thing, and text is much more explored and optimized than audio so lets just give it time
I really appreciate you covering all the AI subjects, thank you so much !
My pleasure!
so this is what i get when i tried using this:
me: "can you see my screen?"
AI: "I can't see your screen as I am a language model and don't have the ability to see or interact with your screen. Is there anything else I can help you with?"
I'm guessing you were just using the ordinary Gemini. Did you choose whether to share your screen or "Show Gemini"? Nah
I'm jealous of younger people who can use tools like this to learn like never before, unlike me back in my day only having textbooks, youtube didn't even exist when i was in school, and if i got stuck on something and textbook wasn't helping me, i was literally stuck.
true. learning is way more accessible now
Same. Back then I solved it the simplest way: I didn't learn, lol.
I eventually quit school and learned more then than with teachers.
Knowledge is pretty useless without the certification from School but still, it's another proof of how outdated the scholarized system is.
29:00 There's a dedicated free app (Be My Eyes) which connects real people to the visually impaired, and is currently bringing in AI like this
interesting. thanks for sharing
I love how the AI gets the perimeter question wrong, it's D)42, they missed the 6 from the short side. AI will always confidentially declare the answer, even if it's wrong. I love AI, but this is the part that is scary to me.
Thanks for this. I hope it’s true. Can you demonstrate using it to setup and build a Shopify store… fast and effective to ramp up SEO and monetization? This might be the AI democratization we hear so much about “Prompt my computer”.
It doesn't work for me, when I share my screen, it keeps saying it can't see the video and it is a language model without video sharing capabilities. Even though the video is already in the chat log screen, it still can't see it for some odd reason.
Finally something useful
I always hate the idea of artificial intelligence doing all our work but this time it has become a tool to improve our skills and solve the problem of difficulty in understanding
I can see myself studying something new and difficult
Imagine yourself holding a book and wearing AR glasses and asking the AI how can I start reading this book and make a plan
I hope that the api stays free for developers forever ! This is amazing im able to make so many good stuff for my personal work and stuff. This is insanely smarter than gemini 1.0 pro api
Insanity, I can't wait for it be packaged into a megamam model
Making megamam battle network a reality 😂
What a time to be alive!
Thank you for going over this. I look forward to testing the upgraded voice options when they are released next year.
The real time conversation option with Gemini 2.0 Flash Experimental seems to error out after chatting for a few minutes, at least it did for me with a dozen attempts.
6:15 This is like having a Professor right by your side guiding you through a project or lesson, and who you can question whenever you please to expound on difficult topics. This is clearly a game-changer. 😎💯💪🏾👍🏾 12:02 Is this for real or are they faking it? Because if Gemini can interpret medical tests and make legitimate medical diagnosis then that's very impressive. I'm lost for words.
26:52 & 27:05 both answers are wrong. And the second one even assuming first would be correct.
Actually the second one is correct according to numbers listed. Anyway that is not correct solution for the assignment.
Thank you, now I will try to create step by step an ideal 3D visualization in 3D max + corona render. As always, I love your channel
Gemini 2.0 can also count the number of coins on carrom board.
Carrom is that pool like game played with fingers instead of cues?
Which is the difference between Gemini live with video and chat gpt4o with video and voice too?
Payment 💀
OMG...."this changes everything".....again 😂
tbh this time it really changes a lot :D
Wonderful.❤. The learning revolution is coming
yes!
Now we just need to update humans so they actually want to learn 😂
Google has obviously been sandbagging until the other players had played all of their best cards.
Now all eyes will be on them as the gain all of the market share in AI products just before they become flush with cash after being forced to sell chromium and android before the end of the summer.
My thought about gemini 1 was really bad, never understood what I was telling it. But this is great to see.
Great video and very promising for the future indeed. I have tried it as well. Just keep in mind. "Grass is always greener"
it's really crazy how nobody is talking about the book The Hidden Path to Manifesting Financial Power, it changed my life
Thanks for the review... Any concerns on security from passwords and confidential dsta when sharing the screen? What about Keeping projects/conversations memory?
I just want to prevent companies from training on the data I input or use it in any way and tbh no matter Google says, I wouldn't trust that they won't use my data. Their reputation screams.
well i guess thats a concern, people that manipulate that kind of information should wait to the point when these kind of models are available for anyone to download and run locally and modify them all they want, until then, the information you share reaches someone's servers
Google has been in the Ai game before open ai. They made the frame work for what open ai and other use today.
true. most of the breakthrough ai papers came from google
good effort in making examples
I wish Google Gemini 2.0 could be integrated with UA-cam soon.
I showed Gemini live a song I listened to on my TV and it correctly identified it. I asked it to play similar songs on youtube. It could not, because Gemini 2.0 is not integrated with UA-cam yet. That would be powerful to put Gemini and UA-cam together. Just imagine all the possible use cases.
thx for sharing. could you share screen and then go to youtube?
Another Hype! this thing just close chat for no reason after 2 minutes, it does not do searches, it does not edit, another hype video thank you
It's great to see that you are in Hong Kong, let's catch up for some AI talk if you come to shenzhen, or if I go to hong kong
Is there an AI voice speaking on this channel? It sounds so strange when every other word is emphasized and there is a short pause between all the words.
Just look when he/she is trying cross the road, and you will get the clue..
you got those images from a blog? something the model might have easily accessed in training I guess.
Fascinating how google avoids using Spanish in their voices despite it's second most spoken language in the US and 500 million speakers worldwide.
New subscriber here from Hong Kong, cheers.
Imagine a humanoid robot with these capabilities!
Would be nice for ai to watch video with you especially learning videos to pause and discuss
great idea! you can already do this with the share screen feature
im pretty sure they are working on something like this. my dream would be this ai implemented in new tvs, watching movie or tv show pausing for second to live ask ai what is that actor how old is he, or where was that scene filmed, what brand car is that etc etc, i spent tons of time doing that on internet, live chat would be so cool
@@theAIsearch i used it to react to me playing skyrim in realtime and every 5 minutes or so it stopped and said "something went wrong" i wonder if its just because they just released it and its experimental, but also they give you less control when you hit the show option in Stream Realtime option. Also in my particular case the AI was too explainy with everything it saw on screen, so it would again be amazing to have some more control of it, i hope they add that
I NEED IT - sponge bob
need to comeback on 23rd now
Google is going crazy
Thier XR ar glasses were amazing tooo
LOVE FROM PUNE MAHARASHTRA🇮🇳🇮🇳🇮🇳 🥰🧡👍❤🤝😍😘😎😃❤🥰
google really has unlimited potential at this point, they are making everything from chat bot, to live chat, to video chat, various ai video and picture generators, not to mention they have their own smartphones, earbuds, smartwatches, imagine this all implemented and optimisedin google pixel ecosystem? they have google powered search that is still top even tho it is infested wit ads and ai stuff but still it THE search engine, add youtube , yt music etc... google might take over everything soon, asking your smartwatch or earbuds to help you do anything in your life, from learning ,teaching , working to simple things at home... this is real main competitor to chat gpt imo, unlimited potential if they get it right. even now its super impressive.
with only this, *not 'only' actually, but you know for 'one' new AI model, Google literally crushed OpenAI (with their not so new 12 days live program haha)
Who let Google cook?🔥
High quality video! Thanks !!!
you're welcome!
Have you seen if the use of a webcam or the view of your desktop functionality is available programmatically through an api? I would love to incorporate this functionality in my application.
Have you tried asking Gemini for a suitable application based on the programming tool or language you want to use?
asterisk sound of gentle wind chimes
I admit that I was a skeptic, but it does increasingly seem like there is no wall. All of these companies keep moving forward. Sure, there is a leveling off in classical LLM transformer architecture with scaling, but overall, AI continues to race ahead with clever new methods. We do seem to be approaching something like the sci-fi AIs at this point. Unclear still if we can hit an AGI that truly has novel ideas and can advance science forward on it's own, but as a tool to efficiently access all the knowledge of humanity, it seems like we are there already.
Can you imagine this Tech on a camera device made for the blind????
I can’t wait for a FOSS version
Thanks, very useful and practical overview.
You're welcome!
I'm totally blind and will always trust my ears to know when a light is green. This has all kinds of potential navigation uses I'm eager to test but you picked about the worst possible one. Even if it's seeing everything clearly, even if you have a good connection there will be just a little lag and, ya know, talking to it while I need to be listening to my traffic might just get me killed. Great idea, awful example.
Thanks for the Hongkong images. It's been a while since I've visited.
Imagine using this with blender 😳
thanks champ. how do you enable sharing of mobile phone screen?
Is the responses really that fast or it's edited?
it's very fast. only 1-2s of delay. try it out!
Edited? Hell no!
@@theAIsearchcant use image generation or screenshare why?
Thanks you so much brother
I'm waiting for the next update for this only on your channel 😉
you're welcome!
Two months ago i been to see doctor because i droped something heavy in the hand, i only go to doctor if i realy feel that i have to. so doctor at the hospital did x ray of my thumb and his diagnosis was that there is nothing wrong with bone no fractures nothing and everything is ok. today after two months i still know there is something wrong with my thumb cause i feel pain inside. so i uploaded image to gemini and there we go it said ,,it apears to show a fracture of the distal phalanx of the thumb. specifically there seems to be transverse fracture line visible across the midshaft of distal phalanx.
Im very angry that doctor did not see that but gemini saw it straight away. Thats crazy
wow, crazy story. thanks for sharing!
For the mobile version, is there a app version for it or do we just use the browser version as well?
I was teaching Google voice AI how to laugh and it broke down as a result. But when I made a new voice chat, and asked it to laugh for me, it laughed properly whereas before, it didn't work. Freaky...
Is it possible to use cusomizable voice model with this?
I want a voice assistant like that, but running offline on my machine, i don't trust google enough to be talking with their AI on my microphone
ok this is fking revolution
It really is, revolution in learning, no? Schools are perhaps no longer needed. It used to be " yeah schools won't even be needed once it's advanced enough" WELL NOW WHAT? This is so crazy
Yes, I can hear you, Clem Fandango
I like yours videos very much 🏆👌👍
Wow nice! Crossing the street on Hong Kong island
it will be awesome when it actually usable, no shade, but honestly can't use it more than a few minutes before it quits and I have to start all over, unable to train either afaik.
Thank you AI search!
You're welcome!
IMPRESSIVE.
조금만 더 가면 AI에게 인간의 교사를 맡길 수 있을 것 같습니다.
why when i ask it to do a search it does not?? dude are you sure we are using the same ai studio?