Honest Google IO Review | Open AI Takes the W, but veo AI Video is dope!

Google introduces advanced generative media tools, including Imagen 3 for improved photorealistic image generation and Vio for high-quality 1080p video generation, soon available to select creators. Additionally, Gemini 1.5 Pro offers enhanced capabilities in Google Workspace and NotebookLM, with new AI features like personalized science discussions and improved email summarization
▼ Link(s) From Today’s Video:
Google Vio: deepmind.google/technologies/...
Google imagen 3: deepmind.google/technologies/...
Lior's post: / 1790456047288631736
Google on AI Agents: / 1790484696557596866
Igor's post: / 1790791005521727959
Google IO keynote: • Google Keynote (Google...
► MattVidPro Discord: / discord
► Follow Me on Twitter: / mattvidpro
► Buy me a Coffee! buymeacoffee.com/mattvidpro
-------------------------------------------------
▼ Extra Links of Interest:
AI LINKS MASTER LIST: www.futurepedia.io/
General AI Playlist: • General MattVidPro AI ...
AI I use to edit videos: www.descript.com/?lmref=nA4fDg
Instagram: mattvidpro
Tiktok: tiktok.com/@mattvidpro
Second Channel: / @matt_pie
-------------------------------------------------
Thanks for watching Matt Video Productions! I make all sorts of videos here on KZbin! Technology, Tutorials, and Reviews! Enjoy Your stay here, and subscribe!
All Suggestions, Thoughts And Comments Are Greatly Appreciated… Because I Actually Read Them.
-------------------------------------------------
► Business Contact: [e-mail]

[@ImTheMan725]

The fact that we can't use almost anything Google showed us, and we will be able to use GPT real time voice by next week. Tells me Google is way behind OpenAi

[@DJ-Illuminate]

Googles opening was so juvenile I literally could not watch it. I think Google has really lost it.

[@wolfbrigade1]

I'm glad someone else noticed this too. Their A.I. hyping is most likely half real hot air.

[@brexitgreens]

By the next week? Source? Still none for me on ChatGPT Plus.

[@mikiex]

Also expect 5 or 6 name changes

[@ich3601]

I thing they fell behind when they over aligned their model. Correcting this took much ressources I guess.

[@starblaiz1986]

Google showed us how they are going to integrate AI into their current products, which is neat. OpenAI gave us a black swan moment with their AI-first approach. As an older mellenial who was there for it, I can tell you that OpenAI's demonstration had very much the same vibe as the original iPhone announcement. People at the time didn't fully grasp the significance of it, but years later they looked back and wondered how they could have missed it. OpenAI had that same moment this week. Smooth, natural interaction like they demonstrated along with all the multi-modality being integrated is HUGE, and right now people don't seem to really be giving it the value they really should. I guarentee five years from now we'll all look back at this being the week that changed everything, and we'll wonder why people weren't more excited about it 😅

[@brexitgreens]

I still can't see anything incredible about the first iPhone. The concept of a handheld touch-screen device is decades old - and a useful realisation has only required good touch screens and good batteries. On contrary, iPhone was overdue and inevitable.

[@drizel462]

You hit it. Legacy tech trying to figure out how to integrate AI into their existing business model while the next generation tech is building new business models with AI at the core. I don't think Google has the vision or creativity to lead this new paradigm.

[@ich3601]

​@@brexitgreensIt was the integration of a touch device a la Newton and a phone with the business model of NTT iMode that made the difference. Apple never invented anything themselve but fitted IT into a nice product.

[@Mercutio111]

Very true. OpenAI (especially Sam Altman) has a compass how to handle these new wealth of possibilities, where Google seems to be fully overstrained and overstressed by it and furthermore terrified by the advancements of the competitors.

[@narottamzakheim5051]

I compare this to the release of the first blue and white iMac. That is what put a pc in every home. Before this computers were ugly grey boxes. The only people who had their own computer were considered nerds. That iMac made computers exciting… sexy… fashionable. Then soon everyone had a computer then laptop which were eventually supplanted by smart phones. I see gpt4o having the same impact in the field of AI. Esp once it integrates into iPhone… The flood doors will open.

[@DisturbedNeo]

OpenAI is going to release a Windows version of the desktop app "later this year", but for the time being it's going to be MacOS exclusive, which sucks because the use case of being able to share your screen with it to show it exactly what you're doing in real time and speak about what's on the screen in front of both you and the AI, is freaking mind-blowing.

[@milorad9301]

Partnership with Apple is coming on WWDC.

[@TPCDAZ]

@@milorad9301 Microsoft put $10 BILLION into OpenAI, I'd be kinda pissed after doing that not to get the desktop app atleast the same time as Mac OS

[@lighteningrod36]

Its about control, the numbers using it. Infrastructure is king. I think they have better models, only waiting for new GPU's and LPU's.

[@nathanbanks2354]

Looking forward to them adding Wayland support....

[@Scorchio-]

Microsoft don't need OpenAI upstaging CoPilot too much. I would expect to see the Windows OpenAI app some time after the big Windows CoPilot update.

[@adrianmunevar654]

The Google music choices for its products... 😅🤣

[@Windswept7]

It screams ‘out of touch’

[@emilrogengellschwaner3555]

Google seems to be about 6 months behind OpenAi

[@vainezaiven6677]

At least.

[@craigmiller2411]

Which is surprising since the employ top AI mind Demis Hasabis.

[@rexvonzombie]

That's what I've been saying. 6-8 months, just about.

[@tornyu]

IIRC Jimmy Apples said Sora was being _used_ and tested a whole year before they announced it. If that's true then OpenAI is probably at least a year ahead with video

[@JoeLancaster]

In what way? Seems more like 3 months ahead to me.

[@ALTERNATIVEMIXTAPES]

Google is really good at showing off products that they will never release....or will never release anytime soon. Maybe they will have this out by 2030.

[@JohnSmith762A11B]

They're just blowing smoke in Wall Street's face as they try to figure out how much woke is too much woke in an AI. This is made difficult as they cannot discuss it internally, externally, or really anywhere.

[@bigglyguy8429]

And it will be in 'beta' for 6 months and then closed without warning

[@greenockscatman]

I like how their car video demo just decided not to put flying cars or holograms at all.

[@HUEHUEUHEPony]

The future is actually decent zoning laws so you don't need car

[@visualride-davidcox2243]

As mentioned in a news article that I just read, the demo of how to fix the camera lever from Search returned a horrible solution. Opening the back panel of the camera would expose and ruin the film.

[@MattVidPro]

just realized that LMAO

[@SpeedyCreates]

omg 🤣 thats google for you I guess

[@fredrik241]

That's just LLMs for ya! The future of human knowledge retrieval!

[@justcallmebrian793]

@@MattVidPro LOL, futuristic city, it looks like a city somewhere in Asia

[@narottamzakheim5051]

Google tried to show how capable they are. GPT-4o is more capable and has actual charisma = winner hands down. People want HER... OpenAI delivers

[@JoeLancaster]

In what way is GPT4o more capable? Seems slightly less to me.

[@bigglyguy8429]

@@JoeLancaster found the bot

[@JoeLancaster]

@@bigglyguy8429 it's a genuine question. I think I've asked like 30 people but no-one seems to be able to point to any specific thing...

[@bigglyguy8429]

@@JoeLancaster Video is not as good as Sora, multi-modal is not as fast and smooth as 4o, everything from G is gimped and censored to the point of being broken, it's always in beta and about to be dropped etc etc

[@JoeLancaster]

@@bigglyguy8429 video might be a valid point. Sora does seem to have a slight edge there so far. Multimodal is definitely faster and more capable though, especially with video input. More capacity also.

[@MrTk3435]

This is great !! They are competing for our approval. Thank you 🔥🙏🔥Matt

[@MrJaggy123]

"They probably dropped some bombs in their pants" 🤣

[@MattVidPro]

Lmao

[@JohnSmith762A11B]

Yes, but since it's Google you can be assured they were smart bombs.

[@SuperFynn7]

20:43 The website says that the Windows app will be available later this year

[@MattVidPro]

thank you!

[@SuperFynn7]

Here's what it says exactly: "We're rolling out the macOS app to Plus users starting today, and we will make it more broadly available in the coming weeks. We also plan to launch a Windows version later this year."

[@ernesto.iglesias]

​@@SuperFynn7 They just don't wanna mess with copilots launchings

[@apersonlikeanyother6895]

If AI is interested in search, It means search will be censored and controlled with no transparency and no means of appeal.

[@bigglyguy8429]

Google's search certainly will be, which is why I moved to Brave search a long time ago

[@Stryker-K]

Seeing these companies compete with each other is awesome lol.

[@dalec2274]

your video seems to be the best i have seen of io shorts

[@brainwithani5693]

My prediction: Google Will release a ton of amazing thing and right when we get used to it they will stop supporting them just because ...

[@ZenchantLive]

Love your content! I would reccomend a simple de-esser on your vocal track. Some of the S sounds are sharp. Love the work bro! BTW this is from a sound engineer lol so take it with a grain of salt, may just be a me thing. But if you wanted some tips on vocal compression or cleaning just lmk!

[@blindstreet]

You mean a talk show needs de-essing? I thought since radio broadcast rely heavily on compression, ear piercing highs were their last concern altogether. Perhaps it's better to suggest they use dynamic mics instead.

[@ZenchantLive]

@@blindstreet give em a break bro lmao but tbh I don't know why this isn't a normal thing for youtubers. KZbin should just have it as an option lmao

[@chanpasadopolska]

I think Google has too many models and projects names, I guess it's confusing even for someone that is actively trying to track it all

[@Dragon_Slayer_Ornstein]

It really depends how cherry picked stuff is, we can't really tell from unreleased previews. I will wait for an opensource version.

[@ThoughtSignal]

Finally a Google veo ai generator I might be able to make movie with

[@swolgan3527]

Does anyone know when the video and "her" audio chat will be released????

[@stateportSound_wav]

In the next month or sooner it sounds like, for Pro plan

[@GaryJr530]

On android we already have Gemini as our assistant, replacing the normal Google assistant we've had for YEARS

[@abrown31415]

Most interesting thing for me was the hint at a new version of Google Glass in the Astra demo. That would be huge! Perfect for an AI assistant ...

[@user-ty9ho4ct4k]

Dude. I've been watching this channel for years now and I just realized I wasn't even subscribed. One up for ya. Lol

[@MattVidPro]

THX for subbing :)

[@chanpasadopolska]

Which model will I use? Both of them, for different reasons, along with others.

[@saksham3]

Unable to join the waitlist. Why is India not on the country list? Even Pakistan is there!

[@eff7221]

Just select any other country 🤷🏼

[@stateportSound_wav]

VPN it

[@saksham3]

@stateportSound_wav Have you even seen the waitlist form? Using a VPN won't help while filling out the form. If one can select any country then what is the point of having that field?

[@stateportSound_wav]

@@saksham3 oh nope I didn’t realize that was the issue

[@MagnusItland]

Hopefully they'll tell me when their new features become available, because I may have forgotten by then.

[@marksokolov1865]

As a person, who has a camera I can say that the advice on 12:09 will kill your camera

[@TPCDAZ]

Typical of google "Hey we made this stuff, it's really cool ... no you cant use it go away"

[@MrNobodyX3]

Out of both 4o and Gemini, as a wannabe developer, I see Gemini, being far more useful than 4o simply because of that incredible token window. It basically means I can put whatever information I need to have in there and call back on it rather than reliant on it being in the training.

[@MattVidPro]

If the long context actually works, in my experience it struggles

[@RetiredInThailand]

@@MattVidPro 'struggles' is an understatement ... but 'Context' features (larger windows, RAG, etc) are not the answer to a better AI (end the 'statistical bias' that is the real 'backdoor' that makes AI, controlled by corporations and governments so dangerous. Opensource or Buzz-off. And deal with the hallucinations! (how can you trust an AI that can be 'statistically' controlled so easily to ignore alternative ideas and opinions, and if it doesn't know the real answer can just as easily make something up rather than say "I don't know'?) Crap (may sound harsh, but it's only in the context of the hype it has been given) like SORA are just improvements on integrating AI With tools to automate workflows ... Automating workflows is fine, and useful, but it's not "AI" ... It's an API that developers can use to ask for AI 'assistance', but it's just the same 'hallucinating, statistically compromised' fuzzy-search-V2024 underneath. When there's an 'AI' that regular people can use to create their own workflows without the need for a team of AI prompt engineers, AI Agent developers, model fine tuning specialists, infrastructure engineers, etc, Then wake me up (or at least stop calling everything a 'game changer' when all that's been done is to provide better tooling integration to a flawed AI, and for very specific workflows that we don't even know how well, or even IF the output can be directed or controlled enough for anything but narrowly contained 'artistry'. Some youtuber making sure it can generate the best versions so far of a 'personification' of a lemon is not really groundbreaking, no matter how cute the lemon (sorry, I understand it's necessary for simple explainers, but it's not providing much insight into a product's usefulness to actual artists in real artistic workflows (which are not necessarily the same kind/class/category of workflows that a developer or UI/UX or a youtube talking heads would want ... again, this is not meant as a knock against you personally, but it is basically the nature of many AI news/review channels on KZbin. It's time to start accepting that AI in its current form is not AGI, and probably not even on the same evolutionary tree if we want to be honest about it. But it is still a useful tool, but so is a screwdriver, and if all the excitement was still over the 'groundbreaking development of the Robertson's screwdriver' rather than over creative construction projects that made use of a screwdriver in its creation, then I would suggest at this point it would be better to start concentrating on how the tool can be used, which tools are useful for what purposes, and a lot less on the how many sizes and colors the screwdriver comes in.

[@driftedsun]

@@MattVidPro in mine it doesn't

[@stateportSound_wav]

@@driftedsunI hear it’s not yet for people

[@driftedsun]

@@stateportSound_wav I’ve uploaded two textbooks (one a poor quality PDF it would have had to OCR) and had it generate quizzes on particular chapters and it’s done a great job with no hallucinations or encroachments into other chapters’ materials. I’ve uploaded a novel in Japanese and asked it to generate a summary of the novel, to list and describe each chapter, and asked it specific questions about happenings, minor characters, and themes and it’s done a great job. I haven’t tried it with coding YET but I’m sure it will do a great job. Although it’s tested extremely well on the needle in a haystack test, it’s important to note that that’s not the only relevant quality when working with large contexts. In my experience, it’s done a great job on large summaries, on picking out minute details, but almost more importantly, on recounting and understanding details that aren’t large but also aren’t extremely minute, which I would imagine is applicable when it comes to coding - picking out functions and bits from a large context and understanding then in context.

[@TheSteveTheDragon]

Googles new slogan: Don't get behind.

[@westbingly]

The retention for race car shorts rank high in the long run

[@Vot8]

The lighthouse exist, i been there. Usuhaia Argentina.

[@ccapp3389]

OpenAI delivers. Google promises they’re about to deliver. “Like totally, trust us bro.”

[@Dude_Wassup]

Are you gonna test GPT-4o for a video?

[@MattVidPro]

once it's multimodal features are fully unlocked yes,

[@Dude_Wassup]

@@MattVidPro Okay

[@Klokinator]

OpenAI: Here's the amazing stuff we're releasing for free in a week or two. Google: Here's the amazing stuff you'll have to pay for that we will probably maybe possibly release next year. OpenAI: Our demos are live. Google: Ours, are uh... they're... OpenAI: They're pre-recorded again, aren't they? Google: STOP BULLYING US

[@michmach74]

Tbf I think in this year's Google IO they had live demo booths, since it was a proper big time event. So people got to play with it.

[@kewa_design]

Tbh its Not a week or two, at least Not for free members 😅 because what WE got in Premium now is far far far far away from what they showed to us in the Event. 4o is basically 4 in a Bit faster for now thats it

[@ALTERNATIVEMIXTAPES]

Release next year? More like next decade or not at all with as slow as Google works. They are really good at showing off products that are not real and will never see the light of day.

[@Brax1982]

Klokinator: Here's me making this comment, although it isn't yet "in a week or two".

[@FRareDom]

Finally we have more ai video generation that is somewhat competitive

[@JohnSmith762A11B]

Er, I guess we will once either of them ships.

[@ArnoldVeeman]

I think we’re only interested in any sora competitor 🎉😂

[@FusionDeveloper]

4:45 yes, as you probably know, certain video formats uploaded to youtube, get compressed more or less. If you want to maximize youtube quality, you have to record/generate very specific formats/settings or convert it (hopefully lossless) to that format, before uploading. For people like me, I just upload whatever I recorded with and don't care about quality, but for a company to showcase a video, they should, especially AAA video game trailers and new AI Video trailers ESPECIALLY if it is fast motion, because youtube will DESTROY the quality if you use certain formats, even if its like 8k resolution at 500FPS, it can end up looking like this video (okay I haven't actually tested 8k, but just saying for the sake of argument).

[@ScaryStoriesNYC]

I won't lie either--if that Google video were out today, I'd drop the other ones I currently use. I signed up for the waitlist as soon as I saw the demos.

[@MattVidPro]

I’m with you on that… why not release it today? Or better yet open source? Zuck is relevant BECAUSE of open source

[@NorbertKasko]

6:33 in the OpenAI Sora demo the waves are unrealistic (they moving away from the beach not towards it) so this is actually better. The model got the waves right.

[@Madeintheshade65]

Google has a very long history of announcing software features and dropping it before it makes it to the public

[@FSK2]

Goigle just tries but in the end open ai beats it hard. Ouch.😮

[@levonkenney]

they said a windows app will come later this year but im not sure why they did not put out both apps at the same time

[@videoeditoranimation1714]

Yep. Matt Wolfe pretty much said the same things.

[@IAmFuzo]

Wow…

[@dvdmon]

As a fellow Windows user, OpenAI not making a Windows app is pretty disappointing, but it suggests to me that maybe they are going to be used by Apple for iPhones and you know how Apple wants exclusivity for everything on their platform, so I guess it's not surprising in a way, but it seems sad if true that Apple is yet again preventing anyone who doesn't use their devices to suffer.

[@landonn3362]

This is no different from when the iOS version of the chatGPT app came out before the Android version. Windows will get its own desktop app later this year. It could just be that the devs at Open AI are more familiar with Apple’s programming language. Apple’s not buying them, just (allegedly) entering a partnership, so I doubt they could have exclusivity even if they wanted to, especially since Microsoft owns half the company

[@stateportSound_wav]

it’s confirmed coming later

[@MoDs_3]

That's need for speed nitro or most wanted!

[@wizards-themagicalconcert5048]

I think you missed the point about the car video, "volumetric lighting' this is a big thing for that video!

[@lpanebr]

What about a Linux openai client?😢

[@brexitgreens]

An open-source client would require captcha to prevent abuse by bots.

[@nathanbanks2354]

It'll probably come because lots of developers use Linux. I'm not holding my breath for wayland support, though.

[@stateportSound_wav]

Would Wine do the trick?

[@nathanbanks2354]

@@stateportSound_wav Wine is pretty good because Steam has spent so much time improving it. The missing part is likely the screen capture. Otherwise we might as well use chatgpt in a browser--slack & discord barely look different in a browser than in the desktop app. I usually use wl-copy for this, but this is wayland specific. If they AI is able to control the mouse/keyboard, this is probably also significant; wayland doesn't have an equivalent to xdotool for security reasons.

[@stateportSound_wav]

@@nathanbanks2354 nice to know, thank you !🙏

[@jp.girardi]

A multimodal model in the same level as sonnet, available in the whole world, and less than half of Haiku? Awesome

[@gabrielmalek7575]

Ive had Google AI search enabled for what feels like years, this isnt a new feature you could enable it in google experiments forever ago

[@googleSux]

Google showing off stuff that nobody has access to competing with Sora which also Nobody has access to. Vaporware chasing other vaporware. Brownlee said it perfectly. They are selling us stuff that might eventually be made public.

[@Mangolite]

Google has a stable organization but is struggling with its AI capabilities. While Open AI has internal organization issues, it far better exceeds natural nuances with its AI integration. As an Apple user, I am surprised that Open AI featured Apple Products throughout their demonstrations. I hope Apple will partner with Open AI for its upcoming AI-based devices. This will make Apple be in the current AI challenge rather than be the odd man out.

[@NOPEME]

I feel like you have not seen the whole google I/O, basically there are some things which you hypothetically interpreting in wrong way, one of them is ask problem with video.... Its more like open google lens shoot the video while talking about the problem then google will search directly just like gpt's latest model.....

[@MattVidPro]

I showed both demos of those two similar technologies in this video, for both of Google's demos I feel that Open AI's were a bit more impressive

[@DaggaRage]

having a separate client like open ai is planning makes more sense for multitasking, working cross multiple apps, who like to just work in the browser anyways? so unefficient. in the long run integrating an ai assistent in the OS is the only way forward. sry for typo….ipad touch keyboard

[@Terny228]

Reverse the Google announcement compared to the Astonishing Open AI demo’s , And I still will never willingly use anything Google AI, Ever!

[@Chuck_Hooks]

I am stoked for AI reading my emails, especially any associated with my politics. I know that knowledge could never be used against me.

[@MaJetiGizzle]

Ilya’s name is pronounced ill-ee-ya.

[@brexitgreens]

That would be Илия (not a real name) rather than Илья.

[@stateportSound_wav]

i thought it was without the “ee” part, more of a two-syllable name

[@king2178]

Gemini can't recall the beginning of a conversation after 3-4 prompts. I literally dealt with this yesterday. It's so underwhelming.

[@xknucklex8339]

just tell me when ima get my copy of `her` app

[@davidpurple3698]

Thanks a lot. Switch to an Apple computer :)

[@dwainmorris7854]

Come on Matt how about videos on drawing and sculpting with AI ? Not just prompting

[@milorad9301]

Is Google Veo trained on KZbin videos?

[@bloxyman22]

Love how they are showing off things like imagen 3 when they dont even let me use the first iterations for it. They are so far behind competition and does not even let people use it.

[@nathanbanks2354]

Google's video generator looks very very upscaled. However if they actually give people access, it could be much much more useful than Sora.

[@HUEHUEUHEPony]

They will release it 1 week after sora LMAO

[@HUEHUEUHEPony]

And then remove it because nobody used it

[@GaryJr530]

🍋

[@retrofuturism]

Veo is spelled veo

[@karenreddy]

I think combining all the tech together, Google did better. But OpenAI beats them at video and the core multimodal functionality in a single app.

[@Mooooty]

Welcome to 1999 Google 😂🤣😅

[@SK-gc7xv]

Of course it is. Google is doing some things OpenAI aren't doing, does some things better they both do, and others - like Poe, Perplexity, Ideogram, etc. - do some things better than either of them. Voice is still very niche, but OpenAI is the leader there. Poe is the best for regular prompting, with bots almost as good as the GPT store and being able to hop between models in the same context. Midjourney and Ideogram are better than any of the bigs where they're focused. This race is hardly out of the gate.

[@fast_harmonic_psychedelic]

Nice! reminds me of PS3 vs Xbox 360 lol.. google is xbox360..

[@bosthebozo5273]

I hope Google goes down like Yahoo

[@SSLCLIPS-TV]

Meta video AI will be the winner 🏆

[@markmuller7962]

Sounds like fanboyism

[@teawa_]

OpenAI really made this google event less exciting. I can’t even hate because this just makes the competition more interesting

[@JoeLancaster]

OpenAI took the W for optics. But it seems to me that Google just has the superior product...

[@tonythereader]

Do you have any intention of testing ALL image and video models for prompt accuracy in the future or just stick to generic benchmark tests? I understand if you just want to stick to the same "I don't care about it being as objective as possible" youtube tests because of the algorithm. I would like at least an acknowledgement of their past mistakes.

[@FusionDeveloper]

8:29 ... 8:32 ? 8:43 hahahaha oh my gosh, this is horrible! I don't know what this is about, I can't get past the music.

[@danielisflying]

Google lost?? They have superior scale to OpenAI and are releasing a lot more tools.

[@nathanbanks2354]

They're behind on Gemini & video, but ahead on TPU's & AlphaFold.

[@ai.mindful]

The keynote was much of Animations not actual market-ready Features

[@chrisaguilera1564]

If this came out last year or even six months ago, I might have been blown away but now after seeing what Open AI has in store for us....well its kinda meh.

[@blindstreet]

I'm starting to hate Google being late to the party despite the claim that they started everything regarding AI in 2017. I thought they invented GPT. Now they are lagging behind everything.

[@povang]

What OpenAI did with SORA is not feasible in real world usage, its too high quality and thus too compute hungry. What that Chinese and Veo is doing is cutting that quality by half in order to serve it to a massive audience.

[@ThanosElohim]

“ILYA” not “iiya”

[@frosty2912]

The videofx created a less then 5 second clip, the search stuff perplexity has been doing for over a year, the Video prompts looked terrible in quality. All of these were cherry picked as the best it can do. Even the demo with the girl walking around with the camera, the ai was really slow and had a robot voice. Googles loosing ground in the AI world and I can't say I'm broken up about it.

[@brianmolele7264]

I’m not even going to bother finishing watching this video because Google only release their stuff only in America and Europe.

[@maragia2009]

Totally missing the point Matt. Google were announcing their products, not one-upmanship against AI. Just doing business N getting the customers excited

[@shujin6600]

Youre slowly but surely becoming an AI boomer. you add scary in front of anything

[@JohnSmith762A11B]

Yeah but it's scary out there man.

[@shujin6600]

@@JohnSmith762A11B there's nothing scary out there atleast for me.

[@Discovery_Nuggets]

Look, I understand that Sora cherry-picked its videos for the public, but still, what google showed not that impressive. So, I would say, nah. The whole presentation by Google looked like Corporate meeting - a lot of talking and zero of showing. At least, presentation from OpenAI showed us how GPT-4o works in the real time. Here is more like: our model can do this and this and that blah-blah-blah

[@Paranormal_Gaming_]

You don't know shit

[@ackerwoman16]

wait cherry picked really?

[@Dave-cg9li]

Yeah, the target audience were clearly Google's investors. This event could have been condensed into 30 minutes 😅

[@MountainViews90]

Its almost like these ai companies don't really know who their customers are yet. They are just making stuff lol

[@marksokolov1865]

Veto is definitely not even close to Sora

[@gizmomismo7071]

Sora is clearly better...actually the chinese sora (I dont remember the name) is also better...

[@seandiamond7983]

You should end every video with "and with that, we're all fkd".

[@FrotLopOfficial]

I read it in Matts voice automatically lmao

[@JohnSmith762A11B]

It would be this generation's equivalent of Edward R. Murrow's, "Good night, and good luck."

[@NOPEME]

they've said AI total 121 times

[@FrotLopOfficial]

No it was 123. AI was said by the CEO once which brought to 121 and then the voiceover says it 2 more times.

[@NOPEME]

@@FrotLopOfficial yeahh😂

[@rionix88]

Most of the google stuff is free, openai is expensive

[@PaulRichardson_Canada]

Google is a disappointment going the way of Blackberry.....

[@JohnSmith762A11B]

Oh please God let it be true!

[@lutaayam]

OpenAI showed me that I’m soon going to have a best friend

[@ronanhughes8506]

The voice Matt uses for his youtube videos is so unpalatable. It's a real killer to his good content.

[@FrotLopOfficial]

I wish they had an option to set search results to "simplest" automatically. Using Gemini on Android is amazing but Gemini just doesnt shut the fuck up. Like when I hands-free say Ok Google, Gemini pops up, I ask "Whats the date 3-5 weeks from today? It answers with lile a whole ass paragraph.... I have to verbally say "and andwer simply with only a single sentence and get straight to the point" every - single - fucking - time. Only then would it say "About mid June". Its so annoying. And I dont understand why these AI's puke info onto us. They could save so much compute and resources by offering simple answers by default and have the option for long answers as an OPTION.

[@blindstreet]

...At least give us the option similar to custom instructions geez.

[@FrotLopOfficial]

@@blindstreet Agreed, thats the first thing I looked for when I activated Gemini for OK Google on Android. You only get the option AFTER making your prompt and you have to physically tap buttons multiple times. I wanna have it automatic and hands free!!

[@ImDatReal30]

Open ai only showed 1 product google showed tons