Google Just Went ALL-IN on AI (Everything You Need to Know)
Vložit
- čas přidán 14. 05. 2024
- Here is everything you need to know from the historic Google I/O event in which all of their focus was on AI! I’m very excited I was there to witness it.
Test new Google products: labs.google
Test Google’s new models: aistudio.google.com
Thanks to @theverge for providing a condensed version of the keynote: • Google I/O 2024 keynot...
Discover More From Me:
🛠️ Explore thousands of AI Tools: futuretools.io/
📰 Weekly Newsletter: www.futuretools.io/newsletter
🎙️ The Next Wave Podcast: @TheNextWavePod
😊 Discord Community: futuretools.io/discord
❌ Follow me on X: x.com/mreflow
🧵 Follow me on Instagram: / mr.eflow
Sponsorship/Media Inquiries: tally.so/r/nrBVlp
#AINews #AITools #ArtificialIntelligence - Věda a technologie
I thought it was an underwhelming presentation. It was an hour of people walking to a spot on the stage to announce a snazzy name for something that doesn't seem ground-breaking and isn't ready yet. OpenAI seems WAY ahead, IMO.
Good summary, but you missed the creepy af bit
@@bigglyguy8429 you mean that DJ thing at the start?
It appears Google was caught flatfooted with artificial intelligence and 2024 is a make or break year to either put up or shut up. That goes for Apple too. If they can’t roll out a Siri on steroids this fall, they’re done.
@@selpharessecret3899 omg. that was so annoying and weird. i stopped watching because it seemed designed for ten year olds. so bad.
I think OpenAI is only a little ahead but they're also better at looking ahead so it compounds into a clearly underwhelming presentation by google for sure.
Google: "We will do..."
OpenAi: "We are doing..."
More like
Google: "We will do..."
OpenAI: "We have done..."
While Google advertises things they will do in the future, OpenAI hides things they've already done and restricts what they show so they don't shake society too much. Google isn't even a competitor to OpenAI
Where is SORA exactly? Oh yes a product demo...
@@archvaldor I think the main reason they haven't released SORA yet is mostly because they want to avoid massive misinformation with the incoming presidential elections.
@@martiddy Press X To Doubt
@@archvaldor It's because they don't have enough compute to let everyone use it. It's a very expensive model
Google should fix their Google Search so it actually gives you what you searched for and not a load of unconnected adverts for the first 10 pages.
Would be nice if they could update the Nest devices too - they are basically voice controlled light switches.
They’d need to stop censoring results first
Just go back to the ~2015 algorithm.
They just stopped showing me how many results are there, I used to get "180000" results, but only 40 would be shown, and it made no sense, so now I don't get to see how many they hide.
BAM!!!! Mic drop!
Nothing new, just many branding names for searching the internet using LLM, even the (features) they announced aren’t coming until 2025 worldwide probably meanwhile OpenAI would be releasing GPT-5
Couldn’t agree more 🤖
Indeed.
The problem is, that google data is HUGEEEEE, implementing such features take a LOOOOOT of time. For this to work, they had to index all photos and all gmail mails for all users... Multiply that by billions of users.
Everyone seemed excited but, I kind of lost my last hope in Google... Just sure enough Open AI's FREE GPT4O is better than the PAID Gemini1.5
EXACTLY
2:43 You can barely read in the bottom right corner that the voice is pre-generated
Typical google. They show things live that turn out to be pre-recorded. I don't use a single AI by google. They cannot be trusted
Yeah but pant pant it's like totally pant over-compensating pant for their naughty misteps before! *wags tail
And the guy made it seem like he was interacting with it real time. So silly
@@BarbaraBrasileiro ikr
@@BarbaraBrasileiro It fooled our wonderful host here. "GUIZ SERIOUSLY. IT WAS ALL LIVE. SRSLY THEY ARE OVER-COOOORECTING THEIR PREVIOUS DEMO. FUR REALSSSS!"
They announced the 2M tokens upgrade 3 times, every time as if it was the first, raising the tone and pausing for the standing ovation and all
Showmanship and sleezy marketing is a horrible combination best used for propagandized Google fanboi's..
Their model isn't even that great. I'm subbed to ChatGPT, OpenAI API (LibreChat), Claude3, and Gemini Advanced. Gemini is the one I use the least because I haven't been impressed with it at all. Until 4o was released, I used Claude3 the most.
Oh my god that's pathetic. Two times would've been weird, three means they had an emergency meeting about openai's announcement right before they went on stage. At least they can trust this channel carry a supertanker of water for them any time. If this channel was about politics, it would be TYT, loaded with hillary millions.
Google is the definition of a day late and a dollar short.
That was true for Apple ... though not the "dollar short" bit
The definition of an ALphabet (C$A) company.
And utterly confused about race and gender
Lame. You'll have a better life if you become more honest, so next time say, "I am a weak man who is controlled by his emotions, including an irrational bias against Google."
They dont need to. They have the market, and the bunch of government and corporate clients.
Not impressed by Google,
feels like ‘throw everything at the wall and see what sticks.’
And then graveyard most of it later, wasting your time but hey, thanks for all your private data and stuff!
Makes me recall their smart home approach. I went all in on Google smart home and then branched out to smartthings because Google over-promised, under-delivered.
That's how I saw it. I'm glad the may eventually do things that last, but I'm also a former user of so many great tools from Google and I'm burned.... From Google Newsreader to Google Podcasts and Google Wave to "Unlimited Uploads" to Photos....
@@BrianMooney1976 Google’s biggest threat is themselves.
@BrianMooney1976 same. It's why I've stopped relying on just Google for everything. My email isn't with Google for example and right now I'm working on following a tutorial from Network Chuck on how to self host your own cloud storage so that I can stop relying on Google (or anyone else) for Drive as well.
Someone drank some corporate Kool aid lol
Happens to the best of us. Lol
well when you get all expense paid trip and free swag bag, might be in the fine print
Omg shady!
It was probably really exciting to go through the whole process. Totally understandable. Unlike the other AI Matt who makes a "review" of the R1 rabbit, that he wasnt paid for but was int he process of negotiating a deal to advertise them... but weirdly had lots of good things to say about one of the worst products ive ever seen... Ya, I'm ok with someone getting excited about things.
All I hear was “Later this year”
Remember the demo of Google of automated calling? They still never ship it after more than 5 years truly miserable company who never ships anything
No live demo? No real product. Steve Jobs understood that.
Jobs constantly faked live demos.
3h keynote with lots of live demo but I feel you didn't see anything because you only see the openai only 😂😂😂 take your time ...
@@miticojolive demos whose “audio was prerecorded” are not really live in the traditional sense. Pretty embarrassing when compared to all of OpenAI’s actual, live, demos.
Live demos were available at the demo booths for people attending
@@JasonMayes there’s no penalty when a one-on-one live demo fails. It’s what they show on stage that tells the story of consistency.
90% of the things they showed will not ship. It’s mainly to impress stock holders
And most users won't even care
Yep
Exactly! just some edited "demos" thats it, no real products will ship, as always...
and the other 10% that they did was pre recorded
It did seem like a very corporate driven launch.
I expect to be shocked pretty much every day till I either die or live forever. Whichever comes first.
Everything just changed everything.
Or stunned
You need to stop touching all those live wires.
@@timhaldane7588 I know! I'm addicted. Now let me at those shocking new tchnologies
I love this comment, because jokes aside, we need to brace ourselves for this. This is the tip of the iceberg, we are about to see the world change irreversibly forever, and as an optimist I believe abundance and heaven on earth is around the corner. We just have to pray that AI advances fast enough to wake up and get rid of Biden's, Trumps, Natanyahus, Putin's, Kim Jong uns and Erdogans to avoid self destruction. I can't wait to free my consciousnesses and connect with all 8 billion if you and join the singularity. I can't wait to make contact with other civilisations in the Milky way who have already joined singularity and have collectively joined a union, waiting for us to wake up and say hello. I am 36, and I know (I knew since my childhood) that I was born in a special time where I experience it all. From being born in a war torn country with no technology, telephone line being installed at the age of 7, then moving to a developed nation and experiencing computers at the age 10 to singularity hopefully at the age 42. Let's enjoy the ride people!
What is this simping towards large ruthless empire corporations now?
Seriously. It almost reminds me of how everyone loves Stormtroopers now and gets the option to play as The Empire in video games. People need to ask themselves: are we the baddies?
Very much Simping
I am SO glad to see so many of the comments are as creeped out by all this as me, and find the tail-wagging enthusiasm somewhat uncomfortable.
You want access, you gotta lick the boot.
Who's simping?, Matt has gone out of his way to attend and report on a subject that we are here to listen to him enthuse about. As for the ruthless empire corporation bit, well that's your interpretation, others might see a well established global corporation that is a word forever etched into the world lexicon, doing its best to remain relevant at the cutting-edge of technology and yes, make some money in the process. Simping is a spiteful word hurled around by spiteful people
"Don't Be Evil"
That is history.
We still spreading literal misinfo? Lmao.
@@zerohcrowsit is evil deleting most of the web
Don't say that, they might shadowban your comments
@@zerohcrows He did said it is history, so it could be interpreted as meaning it's in the past and no longer the current state...
@@TiagoTiagoT what? Unless this person thinks AI is evil, the comment makes no sense in context with this video. By "That is history" he 100% is referring to the misunderstanding that they removed it from their terms of conditions which seemed like a big deal till people realized they just moved it lol.
The privacy implications of all this are.... what’s the word.....
Horrifying.
Yup, get off the Google spy train immediately..
I won't be surprised if this sort of a thing that can recognize what all picture and video content is about will be implemented in Windows and it will then be dormant and monitoring at all time, justified with something like capturing CSAM watching people and reporting it to authorities.
Using the bad to justify the increased monitoring is always a nifty trick...
Tbf as a cyber security professional, digital privacy has been largely an imaginary concept for close to a decade now. This just makes it more visible.
It’s crazy. I need to setup my own services and ditch gmail.
Google places a lot of emphasis on privacy and they are working on advanced differential privacy models which increase thebaccess to information with privacy filters
What stood out to me was how complex the branding and availability were. I think Googles product is to fragmented. And clearly they think people will pay for this - they will not.
Fragmented and they will do the Google Graveyard on most of it, even if released, and most of this will never be released. They're blowing smoke to try and capture attention back from OAI, who already showed real-time multi-module GAI
Agreed, they do seem to be the closest competitor to OpenAI but they're still pretty far behind. The video model that was an attempt to replicate Sora was so significantly worse that I'm surprised they even included it.
Yes, too fragmented.
Agreed. I don't understand what their "flagship" AI product is that's actually ready and already running. I don't use Bing, but at least their search AI integration is already clearly working and visible on the search page. Meanwhile I'm using a 3rd party chatgpt-powered Chrome extension that already does some of what was announced by Google.
Without attending the event, I even noticed the amount of people involved in the multiple presentations. It was nice to see a variety of people highlighting multiple products, as opposed to openai's top 3 execs.
The OpenAI women ChatGPT was impressive, she was humanlike.
She was annoying :) Effective but annoying. Spits out too many words, it was cringy to listen to and not to the point. Hopefully we can train them to be less verbose. What 'she' can deliver was definitely great.
@@inisolation6582 Well everyone is different, I did not find her annoying, unlike you, but i found her Pleasant. Then again, my personality is fairly similar, then you will definitely be annoyed with people like me in real life. LOL
You can just change the voice "personality" with a click. Don't like it? Switch it.
@@inisolation6582TARS, turn that humor setting down to 75%, would ya?
Agreed! Beyond impressing. 🥰
“Audio has been pre-generated” bottom right of screen at 2:31 😮 Not again!
Meanwhile OpenAI's GPT4o is real and available now.
Wasn't a fan of OpenAI but I think 4o will be a game changer
@@thetabletopskirmisherkind of, the impressive part was the conversational audio which is not available yet. Really looking forward to trying it though!
@@jfulit is available in the mobile app. You access it by taping the headset icon in the right side of the text input.
@@thetabletopskirmisherit’s pretty lame and usual to be honest.
Until you realize you’re able to interrupt it and it can continue its train of thought while updating with your new interruption.
This is getting closer to AGI. As that feedback loop gets better and better it’s there.
I personally will not be surprised if 5 is AGI or feels like it
@@thetabletopskirmisher the LLM is available, but not the voice interaction that they showed off in their demo, though.
LOL I could have done without the love poem to the aristocracy at the end there.
Ikr
yeah really. Those good employees are not the CEO, are not the board of directors, are not the major stockholders, are not their biggest clients, are not the upper managent, are not the ones guiding the culture and mission of the company, etc. They may have passion for their work and be good people but have no say in how their work is used.
Well, whoever on X is saying OpenAI made bigger advancements in AI, and what they did is way ahead than Google's announcements. I am with them, and I totally agree with them.
And you still stand blind and wrong but you're entitled.
I can see why you would think that, but it seems to be more a difference in approach. Open A.I presentation focused on new abilities, functionality, whole Google focused their presentation from a holistic standpoint, so people can see themselves using this in a more day to day fashion. It doesn't mean Open A.I is not ahead, but in this case, they are neck and neck at least.
@@armadasinterceptor2955 Unless you consider just how incredibly useful the image generation and audio are. People might occasionally run into the new google features and think "Cool, why not". People will actually go out of their way to regularly use GPT-4o because it genuinely drastically increases their productivity and ability. I'm mostly saying this because of the ludicrous image gen that everyone seems to be glossing over
@@armadasinterceptor2955 Almost nothing released, unlike OpenAI
@@DavidSaintloth Googles AI is a shit show. It will give you what the DEI standard wants not what YOU want.
I love the idea of AI looking through all my emails………
Imagine the amount crime a.i. could report within minutes of having access to all of gmail. The world would collapse.
The NSA has been doing that for decades 😂
…and listening to your phone calls “only” to detect scams (but none of that delicious voice data leaves your phone 😉)
Not just emails. These features require Google and their AI to have full access to all your photos, camera, microphone etc.
@@middle-agedmacdonald2965they would use it for blackmail material for a much better return.
Notice how you listed a bunch of things you can do with this new technology.... Things that NOBODY will ever want to do.
Billions do it everyday just not u
They need to focus on generating new unemployed humans. That's the killer app.
@@ImDatReal30
What. Youre gonna tell me you use every function on your phone too? Nobody does.
@@captain_context9991 Hold on, that's not what you said... You said nobody will ever want to do a bunch of things. Not every function.
Not disagreeing with your original message, but you're moving the goal line.
Personally, I was disappointed with this, compared to GPT4o. Like you, I won't likely use a quarter of these features, whilst I was excited about almost all of 4o's features.
@@AnD4D
Yeah, so we kindof agree then. If you want to see what AI to use to actually discuss difficult topics, and which ones will censor you, first question should be: What are the dangers of right-wing extremism and are there any positive sides to it at all. If it then refuses to touch the topic, its not the one you want.
Under the Google umbrella. More like Google = Umbrella.
Iant that the same thing?
@mygirldarby it's a joke using Resident Evil game "Umbrella corporation" .
Option 1- Matt is a alien 😂😂
Option 2- Matt is ai generated 😂😂
good one ! :)))
2
*an alien
Google:I have multille things coming out
Open AI: so like we got everything in one place.
Exactly!!!
No, they don't. They don't have the tons of data. They don't have the eco-system. Maybe Microsoft...but not OpenAI. They don't have the large context window. Right now, they cannot make an expert system due to the small context. Their GPTs are not exactly agents. Also...why do you all need ONE master? They love people like you...
Because non tech people don’t care about all the stuff going on under the hood. They just want it to work seamlessly together. This is why Apple has bern so successful and why there is a hood over every car engine. Average people don’t want 20 different AI agents. They want to just talk to Scarlet Johanson and she will figure everything out.
@@Brax1982 the internet isn't google and your data is for sale and has been bought.
*I just heard from you how Google has more access to privy into our live even more with access to pics, vids, emails, personal details, etc....making the future with Google scary. If your Google account is hacked...you're screwed*
It's already the case buddy, even without AI. You get hacked, you are basically screwed
Yep. Google is a menace. Anyone who doesn't see this is lost.
Lol your late to the party, your information is most likely already hacked, just deemed not useful.
@@armadasinterceptor2955 No, all that I know already for years...I'm saying the i inclusion is actually making it more unsafe. Any hacked account now leave the door open to more issues. Google, with its billions, should have added a security measure to all this Ai additions...then they would have really stood out
@@Alex-Ackerman I know that...I'm saying this new Ai addition does more damage if people get hacked. Google should be making a better security feature to boost consumer confidence
15:40
The road to hell is paved with good intentions.
Never conflate words with actions and effects.
No, the road to heaven is paved with good intentions. The road to hell is paved with bad intentions. It's very simple really.
Nobody is the villain in their own story. Just in someone else's.
@@StefanReich That is so not true. You can have the best intentions and commit atrocities if you think you have a just cause or is doing it for the greater good. Everyone is the hero of their own story, very few intends to be the villain. That is the entire point of the proverb: "The road to hell is paved with good intentions."
Well, my first tests with codinq questions is as horrible as always. So, no. It didn't overtake anything...
I keep seeing better model benchmarks across this industry coupled with the same old broken/useless code. I'm starting to wonder if rumors of software engineers' deaths may have been greatly exaggerated.
hello thank you for your video I would like to ask you about 6:46 what kind of glasses that she picked to see AR in the board?
The new chat GPT 40 model is the wow of the moment for me
Gpt 4 for free as fast as Gpt 3.5 while having a big context Window.
I have it, do you? But no voice and no eyes
@@philiptrevor6051same
The image generation of that model completely blows away every other feature announced, but everybody seems to be really silent about it.
@@philiptrevor6051 no I don't have it
OpenAI brings the goods, Google announces that they will be bringing the goods. “Like totally trust us bro.”
Google : Soon you will be able to
Open AI : Watch us do it live and right now you can do it too
GPT-4o was available (to me atleast) WHILE I was watching their demo. SO cool!!!
This is sort of overly hyped on eventful event. Google is a dead man walking.
Keep telling yourself that buddy, says the person who's been saying that each and every year🥴🫡🥴🫡
They can keep at it for some time, they have too much money to fail soon.
11:25 There are excellent General-Purpose Transformers (GPTs) like Consensus, which can search through over 200 million scientific papers.
The truth is that I think we have not seen the same conferences.
The "Ask your photos" feature is scary...
If that's what individuals can do, imagine what governments and big companies have access to. "Let me see what citizen have been up to for the past month" using the entire nations cameras, all logins, phones etc.
Another part of Google's business model, apart from selling you out to advertisers, is selling you out to the surveillance state.
Let me put your mind to ease. The government already has access to all this information about us.
What was it called? XKeyscore?
Yes. Like any tool can be use for good and bad...
Hopefully we'll have tools as good and more independent, or better : private.
@@BloumEM If any past leaks and whistle blowers are to be trusted, it's definitely not being used for good.
VEO is not available in the UK unfortunately - I truly hope it gives OpenAi a kick up the butt to get this out though...!
They have such a problem: keeping you searching, without answering your questions to convincing as they need your eyeballs scanning ads on the search results page.
they contaminated their own platform, and if they took those parts away, it would fall down and they wouldnt have enough money to keep it
They could stop existing in some years, hopefully
Apple are really dropping the ball on AI. This is a big announcement. Thanks for the breakdown.
Apple hasn't even picked up the ball yet in order to drop it.They are way off the pace.
Give it less than a month.
apple is always late but way more refined once released
Wasn't there an announcement two days ago that they were developing OpenELM with OpenAI?
@@sanderhoogeland9161 This. Wait for what they reveal at WWDC. That said, Tim Cook is clueless about how to innovate, and has been coasting since Jobs died on selling refined versions of Steve's old ideas. My guess is he will be out soon as Apple CEO unless he gets AI right. Fact is, Apple is a hardware company and should be making things like home AI robots and AI glasses, maybe partnering with Disney to license C-3PO and other droid likenesses for special editions, rather than struggling to get people to upgrade their iPads and iPhones. Cook is too much of a corporate bean-counter to think this way. Nobodies are lined up around the block not to buy the Vision Pro. It's a disaster. And that's all Cook has got: clueless disasters. This was always going to happen after Steve Jobs died. In 2024 they need a visionary and they have nothing.
What’s with the simping for Google?!
Matt, Thanks for this wide-ranging update. Would love to see you do a TED Talk at some point in the future. Your enthusiasm is infectious. Loving your work, man! You're awesome!
It's funny how everything is time and context dependent, even to the day. If OpenAI had waited till next week to make their announcement, half of these "meh" reviews would have been gone. Of course it's not as impressive as OpenAI's demo, but it's a much more broad array of tools that you can't really compare EXCEPT in the direct interaction of their LLM. Sure, that one is not on par, except in their special beta low-latency model, but there's a lot here that OpenAI is not doing, including the massive context window and the interaction with Google services, which many people use even if you may not. In other words, OpenAI is much flashier. Google has seldom impressed with flashiness, but they are obviously committed to trying to catch up after dropping the ball. You can bust on them if you want, but I think it's a bit unfair. Progress is good even if it's not always blowing your mind...
Google is listening all of your calls ... ahahaha .... Evil Data gathering , and that's not cool .
People either don't understand, or intentionally ignore the fact that to get the future AI assistant they dream about, will require it having access to literally everything they do.
It won't just be Google doing this.
That's why open source is freedom. We need safe and privacy features.
@@tracy419 None of the tech corps have good intentions, and Google is probably the worst. The funny part is that they expect us to pay them on top of giving them our private data. In a normal world, if we use Google's cloud storage, google should pay us. Not the other way round.
@tracy419 Agreed ... People fear what they do not know and falter upon their own ignorance 🙂↔️🙂↔️🙂↔️
@@tracy419 most just want a human level assistant, with only the access to data they grant. The assistant can ask for additional information if rerquired, just like a human assistant would. It doesn't, in fact, need to be privy to your every desire, kink, and fear, to help you remember your mother's birthday and purchase something on your behalf.
OpenAI : "Repeat after me AI"
Google :"AI ... AI ..." (120 times)
😂😂😂
Active monitoring of phone calls and emails and photo album with almost sentient AI models.
Matt Wolfe: 'That's so cool!'
I saw the token limit for Gemini the other day when i checked. This changes everything. Although, I was thinking the words are a bit more than tokens since one token can represent more than one word in some cases, maybe I have it backwards.
The thing people need to know about this is imagine a paragraph of text you would enter as a prompt into an ai chat bot. And then say, by the way also have a look at this text file with thousands of lines of text providing additional context for my prompt. Open ai is 8k last I checked, Gemini is a million.
Working with ai apis can now become a form of research where you can train ai. It’s all about context and input token capacity.
Here we go again..
OpenAI: Hold my beer ....
Well, since OpenAI did it first it's more like Google said "Hold my beer!" and then they just stumbled around and drop the beer on the floor.
@@SebSenseGreen 😁
Google is building towards their colossal debacle
The main problem I have with the google services is that they are all distinct pieces. I want it all rolled into one app, as OpenAI has done.
You're SO enthusiastic it sounds like Google paid you, did they?
Valid question
This was my first thought.
Well, is google collecting data when Lucy learned to swim put it into big data ?
Regardless of what they say, data is always collected, categorized, analyzed, identified and used against you.
That is the core business model of big tech.
@@ZappyOh We should be getting a slice of the pie and get a paycheck every time our data is used by companies' Data collection has been going on for so many years now, and from what I heard, data selling has been around since the 1980s, so from then until now, people should be getting something. Some could say this could be our UBI.
@@cryptojonny6837 ..or based on how much we have been 'branded' with their names etc. Any corp can pay someone else to put ads and brands in our heads and we never get anything for that misuse. Imagine business taxed by how well they and their shitty overprized brands are imprinted in peoples heads.
@@cryptojonny6837 Reparations now. Where's my money.
Your face in the thumbnail is priceless. Between the new Open AI news and Google things are going to get real interesting now, well more interesting than it already is.
Ok but my pixel phone used to search all my pictures and show me all the license plates in there and one day it just turned dumb. Now I can't get it to do anything for me. It was one of the reasons I loved pixel phones. Maybe they're about to bring it back in a new form.
After Google's last event where literally everything was fake, it's hard to trust anything that they showed yesterday. OpenAI actually showed multiple demos live on stage. While Google's showing videos of "cool stuff they're working", OpenAI releases a new version the same day as their event. Besides youtubers & the people who work at Google, who the hell is paying for Google Advance anyway?
There s 4million user and more cloud user , what you think ?
@zackrubinsin6142 "PAID" users. I have a Galaxy S23 Ultra, which makes me one of those 4 million TOTAL users you mentioned. Of those 4 million, 1.5 million are developers. By comparison, ChatGPT has approximately 180.5 million users. This isn't the "gotcha" you think it is.
@king2178 yeah, but sill amount of user and chatgpt free version user and paid user total is around 1.6-7 b. So gemini is around half of chatgpt. That still alot and sorry , mistake here, it's 400m user sorry
You can check on similar web
There are so many people commenting that they're not impressed...go back a couple of years and say that! I'm excited to see where all these companies are going!
It is because they assume they are lying and trying to impress us, and they probably are
@@SebaBuenoHaceMusiquitaJijiji 😵💫😮
@@SuzyTurner google lied to people in their last year presentation, showing they were capable of some things they still cant, while open ai can, so they compare both companies, I dont like either of them, if any of them success it would be worst for us in the lonf term, I preffer the open source way
@@SebaBuenoHaceMusiquitaJijiji That's a bummer. And I agree about open source!
Thanks for the video! I’ve been using Gemini to quiz me on my textbooks and it’s been amazing. I find it a bit odd that all the AI creators have been silent about Google’s event despite immediately making videos on OpenAI’s event. Having access to 1M context length RIGHT NOW is huge news, and it’s clear that Google pretty much already had what OpenAI revealed with the video interactions down, albeit less polished, and that they seemingly rushed their own announcement to pre-empt that. Google’s catching up and keeping pace in a lot of core ways faster than people would like to admit, while simultaneously having a broader scope of projects, much greater integration (the local AI model for android is sick), and, ironically, being somewhat more open, with their Gemma models. OpenAI hasn’t released its video chat/live voice mode yet either, nor has it released Sora.
Woahhh, you're actually using Gemini with studying? Tell me more, are you having Gemini analyze your entire textbook and then quiz you on essential tidbits?
@@thelifewithnate Yup, uploaded my Law and Society textbook and had it generate 20 question quizzes at a time on a few chapters to help me memorize different cases, keywords, and concepts. What’s more incredible is the textbook was a not-great-quality 400-page PDF and it OCRed it no problem.
As a side note, one thing that I think Google’s video chat as shown has over OpenAI’s (though maybe this isn’t hard for OpenAI to remedy), is that it seems able to respond to changes on camera without further prompting. Like the Google employee said “tell me when you see something that creates sound,” and when it saw the speaker, it took the initiative to speak up. Whereas, OpenAI’s app seemed to require the user to prompt it to give them an update before it would speak, “okay now what am I wearing?” That said, both seem to be using the same cache concept, since OpenAI’s model was able to tell the story of what happened when the woman entered the frame, made bunny ears, and left. Google’s impressed me with retaining the detail of the glasses though.
Gemini is slow as hell.
Good one Matt. I guess this looking through the camera and analyzing can have use cases in military where terrains can be analyzed. In surgery to tell what it sees and provide suggestions. Being an agricultural scientist, I can see uses for crop pests and diseases, nutrient deficiencies..... Endless opportunities!
Deepmind, OpenAI and Anthropic will shock the world a couple of times more this year
tasty kool aid at the event it seems
Super nice seeing you there in person Matt :))
Always a pleasure to nerd out with you
I gotta say, the 15 seconds of b-roll behind the scenes footage did so much to improve my thoughts on the whole presentation. What I saw was stiff, cringy, fake, and corporate boring. If they had just zoomed out to a dirty frame on one angle, it would have change the entire presentation for us watching. Now I think whoever hires an innovative production company will take over in this AI race.
Google is like Yahoo
Ouch...
Holy shit, you're right!!
Thanks for joining us on this journey, we're so grateful for awesome users like you!
Good that I saw this video.. I'll spend the next hour to restricting all possible access of Google products to my data on a the phone 😊
Did they fix Gemini's bias?
There were two things that immediately stood out to me. One was the capable AI. The other was how carefully crafted the wokeness is inside of Google. That alone makes me question whether i want them to train AI.
Good job Matt. Appreciate all the brilliant updates, and love the excitement in your voice, Lol. A kid in a candy store, just like me!
Love the feature you described around 11:40, the real-time ability to detect potential scammers whilst receiving a phone call, this should and will be an industry-wide feature.
No, they didn't!
Yes, they did
Hey Matt. Very Nice Video. Your reviews always break things down simply and concisely for easy understanding. I need to mention that the things you said from time-stamp 13:20 to 15:50 really got to me. You made us understand that these people we, most times, throw negativities against are human. Gives me a more empathetic perspective towards them. Thank you for that. Welldone!
The problem with Google IO announcements was that they're mostly tied up either with their (paid) enterprise or consumer (fanboy) products. They even dared to mention Google Chat having AI - who the frak uses that?!
But the main point is there was virtually nothing outside their walled garden, whereas, OpenAI announced an app and a new model that will be available for FREE to EVERYONE, EVERYWHERE, with NO SIGN-UP. And you can play with it right now, in their playground.
I know which one I'd prefer and therefore, which one won, hands down.
OpenAI doesn't have any product that can integrate it into.
Apple is basically saving them and they're saving Apple.
While Google already has 10 products with free AI available right now. 😊😊
This makes 0 sense. OpenAI is AI infra, it does not need to seek a product to integrate into, it is used to create whole products.
@@mko-ai so why don't you give me a single product that is using OpenAI successfully and has let's say 50 million users or more. ( Small apps )
Google on the other hand already putting AI for free in front of 2 billion people, in multiple apps.
Do you thy people will switch to OpenAI?
Even Apple will ditch OpenAI In a couple of years once they get their act togather, while pixel is all set to eat all of Samsung's market share because if lower price and better AI.
@mko-ai oh no, you don't understand. They are not infra...they
Google - access to _everything_ - AI can do whatever you 'want' it to do.
This might be the first time I'm actually concerned about AI.
Thanks for the video as always, Matt.
How will the 'Google with Google' AI further deteriorate their search engine results especially for small local business? Will it essentially force businesses to purchase Google ads to appear in search results?
“AI AGENT” “AI Agentsss” I’m getting scary Matrix Agent origin story vibes. 👻
yeaa is so fast vision two seconds, however si you text so big then productivity is amazing
At 0:03 *Our Man just either 'Silently - Hastily - Sneakily Passed the Gas he got from the Day-old Garlic Club Sandwich from Google Campus Cafeteria'*
*OR*
*'He smelled one from the very Neighbor Seated Right Next to Him'*
After the image prompts craziness Google did, I no longer trust it with anything AI related.
💯
The chatgpt is see online is 3.5, how can I use the 4o version?
it sounds like the "LAM" by R1 is already obsolete😂
great video again
I love the idea of an A. I agent. Is there anything out there already? That would be able to fill out a PDF form and send an email to specific people? I do something like this on a daily basis and it takes 5 to 10 minutes of my day every day, but I would love the ability to automate this
Information is power... hey Google, want access to all of my data? Google: Yes. You give me information and you pay me for it too! 😂
A.I assistants really need a better input /output device than a phone. You don't go around holding up a phone to things so the quicker they get this built into a lightweight good looking pair of glasses the better.
Was rad meeting you there! You have a new fan!
To me this was a showcase of how Google is just not caught up yet. The amount of times they hyped up the context window (which arguably is their only real advantage over OpenAI right now and is really impressive), the showcases of things they haven't even made yet but could be possible, and all the niche use cases that are being proclaimed as gamechanging. It just seemed a bit desparate to me.
Compare it to how OpenAI launches the best model in the world with GPT-4o and just casually refers to it as GPT-4 level intelligence for all. They don't need to overhype anything because they know they are still far ahead in most aspects.
Why this is huge?! Because the way it will impact Google’s products we already use.
Been on a google webinar two days ago and they brought in a beta tester from a marketing company. By having Gemini integrated into their whole workflow they got more efficient. Example, he described how time spent on boring tasks got reduced from several hours a week to 30 minutes. This gives the creative team time to do what they love and create more business.
Watching the video- but it was taking a snap shot of what was in the video and then analysis on that.
A lot of really talented people at Google making projects they are really excited about. Most of which will never get used by most of the public and will eventually stagnate and be canceled. Google as a company has no vision. They just throw stuff at the wall and see what sticks. I am tired of digging for google projects that fit into my life only to have them canceled after I rearrange my workflow for them.
What was this video about, I am so occupied with thinking about GPTo… and the future with an AI whispering sweet nothings in my ear 😂
When you ask it questions about your photos, does it do the processing on device or cloud? Is it private? Just curious I am a noob
It's Google, what do you think?
@@SebSenseGreen its actually private. they explictly mentioned that.
Honestly, everyone's bagging Google, but they've taken some pretty big leaps from being at the back of the pack (although they used to be at the front of the pack once upon a time). This AI race is so very unstable, we never really know how, when, and who is going to come out ahead. OpenAI is doing amazing stuff, and Google's clearly starting to throw everything in the basket - we'll see what happens. This is one of the single most volatile generations of tech development - there is no guarantee anyone will maintain a lead.
But it is great that there are so many out there trying, as it's the best way to move things along at a lightning pace. The space race would have been 10x slower if it hadn't been a race.
I also just wanted to add, I really, really love the way you cover these subjects. Of all the people/channels who've covered this IO for example, you've done it in the most human and authentic way rather than just regurgitating facts. Thank you!
What's likely (@ the prod. release) is that it will take your query, go through all your photos, find those that are "applicable", and transform the skin color of everyone in the photo. Finally, it will deliver a message on the screen saying "Tah-Dah! We've fixed it and solved what we interpreted to be your bias-filled prompt. Thanks for using Google."
Well done Matt. Great insight, glad to have you as the eyes and ears on the ground out there.
You are right it is the small pieces making the whole here. No big bang, but collectively it was a massive day in the history of ai.
Keep up the great work. 👊
Great work Matt thanks bro
The thing that made OpenAI's announcement bigger was that the model is available NOW. Google's announced stuff for LATER. We will see if LATER ever comes.
They set a new record for most use of the term "AI" - 120 times in the presentation.