Google Strikes Back (The AI Wars)
Вставка
- Опубліковано 27 тра 2024
- Google finally responds to the market pressures from Microsoft and OpenAI. In this episode we take a look at how their AI plans were unveiled during the Google I/O developer conference.
Intro Track: • Burn Water - Nostalgia...
Outro Track: • Burn Water - Soul Mates
ColdFusion Podcast:
/ @throughtheweb
ColdFusion Socials:
/ discord
/ coldfusiontv
/ coldfusion_tv
/ coldfusiontv
Producer: Dagogo Altraide - Наука та технологія
![How We Became the Loneliest Generation [Documentary]](http://i.ytimg.com/vi/I19btmIBhx0/mqdefault.jpg)
Wild times that a UA-cam channel can have a series called "The AI Wars" and it be about real life events.
What a time to be alive..
AI is kind of a buzz more like machine learning wars.
@@monsterhunter445 Machine learning is just a branch of artificial intelligence, so is not incorrect saying AI wars instead.
@@monsterhunter445 Machine learning is AI, the NPC in whatever game you've ever played is AI, the Algorithm in UA-cam or Netflix or Amazon that recommends content for you is AI. AI is a technical term not a magic word describing a super smart machine or even a machine that's near the level of humans.
@@NightHawkExpress A global pandemic struck, Russia invaded its neighbor, a British monarch died, and there's a technological arms race
It's like the 2020s are trying to speedrun the events of the 1900s
That Google had to "strike back" when they had, and sat on, all that technology for years, is telling.
Not really. They revealed Bard and it was rubbish. They've learned from that and polished up the rest of their products so they don't flop the same way.
Think of games that are highly curated and polished when presented for the first time, yet when they are released and can be truly scrutinised they very quickly fall apart. No Man's Sky and Cyberpunk 2077 being prime examples.
@@Xune2000 it’s clear they had the talent and the tools to do this years ago. No way they turned Bard around this quickly. They were probably running slow to handle the ethics of the issue, but we’re forced to remove the brakes.
EDIT: Dagogo made this point in the video already
That is a how a holdings company works.
Anytime a hint of competition comes up. Alphabet inc just uses their access to capital markets to buy out the competition.
Then do nothing with whatever you bought because holdings companies are trash at developing other than monopoly control.
Maybe they rather be known as the safe AI version (populace safe version) rather than being first AI search option. We can use ChatGPT as an example where they have been under a microscope due to users being unable to use it properly and reasonably and because of that, its press has been rather negative. However, by google sitting back, it was able to differentiate itself without simply being the first to the scene and damining its reputation.
@@BunkerSquirrel that's why monopoly is bad for progress. they only started moving faster when they felt threatened.
Giving AI the ability to "fact check" is dystopian AF, especially when these AI are backed by big tech companies.
As an engineer I could not agree more.
Below this is the reply I just gave to a thread started by AlanKey86 that goes _"Within the next 10 videos, ColdFusion will slip in 1 where the script's been written by AI... the test will be if we, the audience, can spot it :D"_
Considering the malignant epidemic of technical ignorance that constantly bedazzles the general public anytime somebody does yet another over-hyped story of _"wow-wee future tech"_ that you all suck down into your vacuous brains then NOBODY should expect many people to notice if the BULLSHlT they are being told comes from a clown or an AI.
Yesterday 40,000 children died because we didn't feed them, today another 40,000 will die because we aren't feeding them and tomorrow another 40,000 will die because we don't have food for them. And that's before we get onto the 10s of 1000s who will die from simple infections and other easily dealt with ailments. MEANWHILE we have a room full of idiots who think being able to photoshop better is worth spending millions of dollars on.
I heard a term recently _"Scientism."_ Its the ideology that science and technology will somehow magically fix everything. As an engineer I don't know which form of technology will magically start feeding children or the magical tech that will fix the climate WHEN M0R0NS keep thinking this crap is more important.
@@tonywilson4713 So... what are you doing about the children?
@@felixisaac I'm trying to help make the agricultural sector produce more food though better land management.
So far I keep hitting walls of ignorance.
WTF are you doing?
@@tonywilson4713 That's cool.
@@felixisaac There is the most terrible aspect of that.
There's nothing you or I could do that will stop the next million children dying of hunger or the other million dying from a lack of basic medicine.
All we can do are those things that might make the world a better place at some point in the future.
I'm actually an aerospace engineer and when they did the first road map for the Australian Space Agency the very 1st item in the list was advanced space based water management with the claim it *MIGHT* be available in the mid 2030s. Its an area of space technology called Earth Observation Sciences. My attitude was _"Why TF are we going to wait until 2035 for somebody to just gives it to us. Lets just get on with it."_
Australia currently has a population of 26 million but produces food for more than 75 million. If we mismanage our water (and we have a long history of doing that) then we don't produce as much food as we should and a lot of people go hungry. However - if you try and tell the Australian Government you want to spend a few hundred million so that it will help starving children in other countries it wont go far. There are these people called economists and they'll ask _"What's the business case for that?"_
So I tried another tack. Australia's ag sector employs over 300,000 people and is worth just $200+ Billion a year. Our tourism industry employs over 550,000 people and is worth over $55+ Billion a year. So collectively we have a lot of people dependent on the management of our rivers, land, forests and ocean areas.
So I pushed that as simple straight forward economic reason - protecting jobs & money.
*I GOT NOWHERE.*
Unfortunately I'd upset a few egos, which is an ever bigger issue than economics when it comes to politics.
What we need is an AI that can give legal advice, better than a lawyer. No more $700 an hour, people need justice not bankruptcy.
Just add an AI Bar license (with standing) and an AI judge (with commensurate powers) and you're all set. I appreciate the spirit behind your post but the blood-sucking lawyers will guarantee their unending blood supply via legislation. It's what they do. I'm kind of an expert because 1) my dad is one and 2) I married one. Oh, and the irony is they hate lawyers more than both of us. Go figger.
I think i does
It nearly passed a law test.
Yep it's coming early 2024 with Chat GPT5.
What we are about to see in the world economy is going to be catastrophic for how many jobs will be lost
GPT4 passed the bar exam and a man has already used it in court as his lawyer.
Within the next 10 videos, ColdFusion will slip in 1 where the script's been written by AI... the test will be if we, the audience, can spot it :D
I think he is already just a guess.
What makes you think that he hasn't done that already?
Thats the most pointless thing i've read all day. Congratulations
For all it is worth, all of them could be written by AI because none of these chlurmcklets know anything about reversible computation. Don't be a chlurmck. Don't be a chlurmcklet.
nah men, by now it is imposible to accurately spot
Would be nice to see AI block all Google's advertisements.
likely, the opposite - it will try every opportunity it can to promote/direct users to services from Google ad network
@Shajirr hmmm. Maybe. If the answer had "sponsored" answers people will drift away from the platform. But if the ad is like banners on the bottom, there will definitely be methods to block them
@@Random_dud31 "If the answer had "sponsored" answers people will drift away from the platform." - but they wouldn't know the answers are sponsored. AI can just convincingly describe /arrange them in a way for sponsored answers to seem like the best option available.
AI language models are conveniently very good at sounding confident, even when its spouting made up stuff or outright lying.
adblock plus does that for you.
What about hacker AI?
As an elder millennial, one of the few advantages is having lived through the Great Recession. My advice. Reduce unnecessary expenses, increase your savings by investing in financial markets and do not sell. One thing I know for sure is that diversifying your income can help insulate you from much of the craziness going on in the world.
That's true, I'm thinking of investing in stocks or digital assets to grow my money for the first time, but I lack the in-depth knowledge and mental toughness to deal with these recurring market conditions. please any advice or pointer on how to outperform the market producing good returns.
“There’s more and more of a concern that incoming data is revealing that the Fed might be a little bit behind the curve than maybe they expected heading into this year,” said Bipan Rai, North America head of FX strategy at CIBC Capital Markets.
Only a good FA will be enough to guide you through the current market volatility. I've been speaking with an advisor for a while now, primarily because I don't have the necessary expertise or stamina to handle these recurrent market conditions. The fact that I made over $220K during this downturn proved that there is more to the market than the typical person is aware of. The greatest course of action right now is to have an investing consultant, especially for people who are nearing retirement.
@@graceocean8323 We’re only just an information away from amassing wealth, I know a lot of folks that made fortunes from the Dotcom crash as well as the 08’ crash and I’ve been looking into similar opportunities in this present market, could this coach that guides you help?
My Portfolio is managed by JEANNE LYNN WOLF, whom I found out about and contacted after seeing her on a CNBC interview. It has since acted as the starting and ending point for the games we have prioritized. If tracking is required, an internet search can be done with her name to reach out.
Honestly the battle between the two companies ultimately just helps out the average person. Although we may make fun of AI responses now, it is pretty crazy to think just a couple years ago we thought voice assistants like Alexa, Siri, etc were the most amount of AI that we would be exposed to.
I've never thought this.
It's pretty silly to think that after Siri/Alexa there would be no progress on the AI front.
However the rate of development right now is damn impressive.
Really? That short sighted?
@@ingmar89 it's much the same sort of progression we witnessed when going from 8-bit computers and video game systems to 16-bit, 32-bit, and so on till we now have graphics virtually indistinguishable from real life, and computing power that was once just a dream.
It helps the average person… until ai advanced enough to make many of their jobs not needed anymore
nidhi
I was chosen to try out an early form of Bard months ago, after about 2 minutes and 5 prompts I laughed and closed it. Impressive to see how far it's come!
Bard still sucks compared to ChatGPT.
It's laughably bad. I asked it can you remember previous conversation? It said yes and we talked about your mother's dementia. But I never had any conversation about it with bard. Chatgpt is at least thousand miles ahead. Bard not only generates faulty response but the same answer over and over again
@@jamesjonnes Bard sucks in general. I guess the sarcasm went over your head lol
@@cpoteet8008 No, I tried using Bard some times and gave up. ChatGPT is superior to Bard in every aspect I've tried. It even told me it could read images, but when I provided links it gave me nonsensical descriptions of the images. That was like, 3 days ago. I simply abandoned Bard for the time being. I'll try it again when the news talks about it again, perhaps by then it will be better. For now, it's just an option in case all other LLMs are inaccessible. The only thing I could say Bard is good at is being fast, but I don't need that at the moment.
I have a number of concerns with Google's AI, but the main one is that if someone is going to send in-depth questions to their software, and it helps "sift through" the information and only gives out the important bits, it leaves the potential for the software to give only one opinion of things. Their opinion. Being able to go through the information ourselves and making our own observations and conclusions is very important, letting the AI do that for us will make us more reliant on AI for our everyday lives, and will seriously hinder critical thinking skills.
this is one of the reasons why they sat on it. Microsoft has the same issue.
Easy fix? Ask the AI and then search it up yourself. 🤷🏽♀️
They already de-rank content they consider harmful in searches, which has gotten more broad over time. That's just standard Google behavior, unfortunately.
Don't ask it for opinions, use it for what it's useful for, which is making text more readable by changing its form.
Turn blocktext into bulletpoints or a table, made an additional summary, show the sources, add information about xyz.
it's not your opinion maker.
And mostly, it will be useful for putting you on the right track about finding an answer. "Hi I need to find a song made by a blond man, i can't seem to figure out who it was, it contained a chrous singing day ey ey" because of the chat function, you can get closer and closer to the answer. "give me a list of blond singers who have a song with the line day ey ey"
What's going to be much more powerful is the AI programming, the fact that it can create apps based on text is going to change things immensely.
There is also a distinction between important questions and inane questions. An ai's opinion on "Where should we go to eat tonight?" probably isn't going to come with life altering ramifications. But it seems like once people start to trust the ai with the easy stuff, they could easily become completely reliant on it for the hard stuff. Abra cadabra; cult of ai!
especially that it is so politically biased, go find out
I honestly think any art, idea, or anything made by ai should be stamped with a label that says it was made by ai like if ai made an artpiece then there should be a little watermark somewhere that says ai generated or something and i think this would help artists because the way i thought of it was you go to the store to buy food but some food is artificial and has gmos and some food is all natural so i feel like people will treat this the same way i fell like ai art wont be premium and it will feel artificial so people will want the "all natural" product, so basically what im saying is ai art should have a "gmo sticker" and that people might value art made by humans more
you're the only youtuber I have the bell notifications on for. really great stuff! thanks for keeping us updated
I used to work in telephony and had to deal with the shortcomings of telephone systems to deal with machines that end up calling machines. So, I used to riff on the old "have your secretary call my secretary" meme by saying "have your machine call my machine." Well, now it's have your AI call my AI.
Have you seen the movie "Her" with Joaquin Phoenix? Looks like this movie was looking less far ahead than most of us thought. I personally have been waiting for this breakthrough of having an personal AI assistant for nearly 30 years, although I didn't know then it would be 30 lol. I then thought it would be 2050.
@@paulmichaelfreedman8334 I did see that movie. I think about it often. At some point, someone needs to do one with the gender roles reversed.
@@TesserId That would be less apealing to the public, furthermore the movie also shows that he is not the only one who started a relationship with his personal AI, like his woman friend, where the roles are reversed. It is revealed that every AI is not a separate entity, but one huge AGI posing as all of them, hence knowing everything. The reasoning is that once one becomes omniscient you are in fact the definition of a God. Knowing everything means being able to do anything possible in our realm. That said, I just thought of another brilliant movie, Transcendence with Johnny Depp. That one really appealed to me.
@@paulmichaelfreedman8334 Thanks, I'll have to check out Transcendence.
I always love watching your material. It's always professionally delivered.
The timing of the accuracy with which you describe such a technical and critical subject is pretty impressive.
Because you are not from AI specifically if I remember.
Honestly, one of the best series out there !
Accuracy isn’t that good. He’s too idealistic
@@rationalistfaith No prediction can be accurate, they talk about something that didn't happen. The stated facts about the AI wherabout are correct, which is impressive if you don't work full time in the field.
Google's "magic eraser" has always bothered me, and I think it's because you can so easily take pictures that don't reflect reality. The demo in the intro shows exactly that.
Also, metadata is not sufficient to address AI-generated content. It can be removed or a screenshot can be taken of the generated image. Therefore, detection must come from analyzing the pixels themselves. Even if this kind of detection is perfect, however, another person will use another AI to detect and remove the subtle signs of the content's AI origin. Pandora's box has been opened, and metadata is like trying to close it with paper towel.
00:33 Move yourself in an photo to seem like you are catching the water
The audience: woooooooooooww *mega aplause*
14:02 Help improve effectiveness of treatment for diseases like cancer
The audience: meeeh *few claps"
indeed, people are dumb. AI will EASILY take over.
Please do as much on AI as you can. The videos are gold!
Okay, admittedly I'm stopping halfway through this video to write a comment, but all of this stuff that Google is showing off is doing little more than making shit work like I already expect it to.
Google is the absolute king of providing crappy search results, and it's not because it's indexing or search algorithms are bad; on the contrary, they tend to be among the best. Google's biggest problem is that it tries to answer its own version of the question that I'm asking, instead of the very specific, carefully-worded search query that I typed into it. And the reason I use it like that is because if I don't, I get absolutely useless results. I have to jump through hoops in order to get Google to understand what I'm actually looking for.
It used to be that you could use operators before words like "+" and "-" but in their infinite wisdom, they eliminated those in favor of a search that "answers questions." Google doesn't want you to feed it inputs, they want you to "talk" to it, like it's a person that understands speech, except it's not and it doesn't, nor do I want to type out an entire sentence when I know that half of the words useless anyway and I want the search engine to focus in on one or two very specific key words. Search operators are powerful shortcuts that help make the search engine provide the best results in the most efficient manner possible,
Google's desire to "answer questions" instead of providing actual search results is killing the thing that makes them all of their money. This demonstration isn't particularly compelling, especially since it's a scripted presentation and we have no idea whether or not this is actually working the way they're demonstrating it. The questions they're asking are very basic and obvious, and it's actually kind of maddening that they're presenting this stuff as if it's a major breakthrough when people have been trying to use search engines like this for decades now.
Couldn't have said this better mysel. Unless I asked AI to do it for me. You're absolutely right
You are spot on. Not only that, but most of the things showed in this video has existed more or less for a decade now. The photo tools have been available through various apps for years. The medical MRI / X-ray scanning by AI has been done since 2016. What I see as most promising is the protein folding project, as well as further improvements in the medical industry. But the biggest worry is Google "training" their users to think the way they want to, through AI search results and recommendations. And imagine how much more data they will be able to process about each individual with a Google account.
I'm so glad I'm not the only one who feels this way, and misses the boolean operators!! This is exactly the kind of feedback Google needs to hear
5 years ago I've never had a "no results found" page.
This year I've encountered it at least 20 times.
Even worse are the search results that magically echo chamber what you type into google and give you nothing while leading you nowhere.
Forum search results are pretty much gone.
95% of EVERYTHING is just pure product placement and "reviews".
Googling anything remotely obscure is a huge hassle now.
Thanks for the great work you make in creating these masterpiece episodes.
I love this AI war series, Things are changing around here now👌👌
09:29 this moment reminds of what was described in the Avogadro Corp book, the first of the series about AI. I recommend reading all 4 of them to see how the emergence of AI turned out for humanity, very interesting indeed.
Your videos never cease to impress! What a wonderful overview.
Grand segment, enjoyed the information given in this piece here. Great job.
This channel introduced me to AI a few videos ago! Keep making videos on this subject please. Cheers
With all this exciting stuff coming out, there is only one thing left to do: Find a better name for the AI than 'Bard' 😂😂
honestly though. you'd think they have access to better names haha
Bard is not just an ai's name
What about Brad Pitt 😂
B.A.R.D. is an acronym, not a name. It stands for Binary Array Recursive Differentiator.
They could have named it "GAIA" - Generative AI Assistant
How much progress in just 20 years
In just 100 years
And 99.99% of today in last 500 years.
It immensely inspiring and fascinating for humanity.
Another great video Dagogo!!! Google will dominate in AI for the following reasons
1. They have heavily invested in AI tech for the last 13 years. Their acquisition of DeepMind provided them with patents and the best AI talent.
2. Google has all of the search data including text, photos, audio and video on every topic. Of course AI is created by learning data.
3. Google continues to attracts the greatest talent in the world. Even after all of these years, Google is still the most desired company to work for in the world.
4. Google has no limitation of capital to invest in AI.
So in summary, Microsoft, Amazon and Facebook can create AI of internal applications but the greatest commercial applications of AI will come from Google.
You've been on a roll pushing out so many great videos lately!
He started using AI ..
@@KoushikPaulliveandletlive 🤣
This feels very much like a google ad.
@@Rivanni It should be disclosed very specifically in my opinion
i love the garage, every coldfusion video wraps me up in good vibes and engaging conversation.
Brilliant as usual. I need to check out the full io event.
I don't want either microsoft or Google to win
I want both of them to suffer from cut throat competition
This way, the users win
nice! +1
Microsoft is gonna strike back surely so hopefully you're right
The users win if open-source projects win the AI race. And according to an internal memo from Google, they likely will
@@shadowdragon3521 wouldn't open sourcing advanced AI be dangerous?
I want them to win but i want them to suffer everyday. The consumers are the ones that win
Id like to see some reporting about how different countries are handling these new AI launches; for example Bard isn't available in Canada and the EU and there isn't accurate reporting about their process
Looks like Developed Nations are finally letting third world Countries advance.
I remember reading about a country that made AI an active member of its parliament. They will see what the AI suggests when it comes to legislating.
bruh what you on about the entire planet is the US
@@d.a.7689 I like Ai but that is the dumbest thing anyone can do. It's one thing to use it for reference, it's another to have it be an active member of legislation.
@@ExValeFor ha...ha...ha?
I would have expected Google to integrate these AI tools with Google Assistant in some form, along with search. Why haven't they? 🤔
The big companies aren't the only players in the AI Wars. It's also the general population with its open source projects. And if AI gets regulated like the big companies want, we're going to be the players who lose first.
Thanks for your content, it fills me with joy knowing people like you are around to entertain and educate us from an objective perspective.
I'm terrified how fast things are changing
If you ever decide to step outside digital world, proceed with caution as you will encounter even more dramatic changes
same here. this is giving me anxiety 😅
It's all hype.
In my opinion, you run the best UA-cam channel. Your topics and the way you narrate them is spectacular. I'm here since the ColdFustion times. Thank you for the great job. And also, let's see where the AI wars is gonna take us.
Announcements felt even better with Coldfusion music in the background
11:50 he mentioned that they can retrain the bass models. That's a good example of how AI is creating new jobs. I'm one of the people who work on mechanical turk training these AI do that means more work for me
I think this is the real AI story. The story where AI can be responsible, improve work flow and assist with reliability of information.
That is the good side. Unfortunately, reality has both sides. There are good and bad people in the world. How individuals use the product for good or bad is associated with personal intentions. Good luck trying to regulate that. But it would be helpful to educate individuals to see the benefits of using the tool in a good way. This would be a monumental task because not everyone sees things the way, and that's by design.
It is likely AI will be used for so much misinformation
Love your AI videos, always makes me excited for the future 😊
Thanks for these very well made and interesting video's ! Congratulations.
Alternate title : Google Event 2023 in 17 minutes
Was waiting for this to drop!!
Exceptional work as always.
An A.I wrote my recent job resignation. It also helped me fix my CV, and to write an application off of my own written application, but making it way more straight to the point and appealing.
Got so many job interviews, and currently in a new job. Still getting job offers.
The AI wars is so interesting to watch youtubers talk about what is happing and great video man :]
I wonder if Bard being integrated into UA-cam will mean that all the spam bots will be detected and dealt with accordingly.
But knowing them they'll probably just crack down on people using naughty words and leave the bots alone
Precisely. Or understand hidden contexts in videos of topics they don't want on their platforms (like swears written on the video to not having it said vocally) that previously would escape the bots, to hurt smaller channels even further. I don't trust these companies a bit in anything they are doing.
AI needs to get rid of bullying others on the comment section too
Also the model can start vetting comments for ones that would be harmful for Google. A comment with a hundred thousand likes and a sceptical outlook on any Google's sketchier behaviour in the future might just vanish into the digital void.
and somehow more targeted ads!
How will it deal with actual bullies and not with "perceived bullies"?
As someone that uses Photoshop on a daily basis, can't say i find that first example with the waterfall an advancement. People that were "oooohhh ahhhhhing" are probably that never touched any modern editing tool except their phones. Photoshops own internal ai with content aware does a great job as well as their neural filters.
Show me that same photo of the woman in the middle of times square in NYC with people in the background and lets see how well it swaps her location and fills in the guessed area. Grass and foliage is so easy to swap and paste.
Cope your job is about to be taken by the silicon valley overlords. Kneel down and put some respect to your new AI overlords.
ONCE AGAIN, BEYOND PAR AND PHENOMENAL DOCUMENTARY.
SIMPLY SUBLIME WORK, DAGOGO!!
BIG SUGGESTION: Please put official links of all AI apps that you mention (in video) in your video descriptions. It's such a hassle to pause the video and go around googling them, would be much better to click thru the links after I've finished watching the entire video. Thanks🙏 And do keep up with the AI series 💙
^
lazy mindset
@@xxTHExxABYSSxx mind your own mindset
@@Eddy-Cool likewise
Do your own research and find links. We all do it every day.
@@xxTHExxABYSSxx As I said, MIND YOUR BUSINESS, MR. WE. you ain't needed.
I need AI to turn instructional videos into actual instructions so I don't have to watch a 12 min video to find the one step I actually need help with. Please make this happen.
12 min means ads at the start + midroll - so why would they do that if thats how they earn money ?
Already being worked on. Honestly theirs probably tutorials out there on how to set something like that up with current ai but I don’t think any big company has truly implemented something like that yet
@@seowebuacus they don't wanna give up the future to open source llms yet
There's a chrome extension that uses ChatGPT to summarize UA-cam videos
I want Ai that i use my own factcheck if the content of the video really match the title ..
Also does factcheck AI will delete or block all videos that AI believe to fake ?
*Fucking scary and fucking exciting at the same time*
Probably the best way to put it. I feel as if we stand on the precipice of something big. How this will all play out is the real question.
If I had to guess, this must have been what people felt when they saw the A-Bomb used for the first time.
I'm glad they're also investigating AI detection and responsible use tools.
Damn @ColdFUsion, you know, I'm really into tech and stuff. And I really am following all these new delevopments. But tbh, watching your vidoes when I'm coming back home, deeply drunk ... these are just the moments I'm really enjoying watching your videos. It's just about the feeling of how unreal all of the stuff is, at leat for the moment being .. Imagine watching one of these kind of videos back in the 2000s, when the first smartphone was leaked ....... mindblowing. Love you :)
That _"400 million years of research in a few weeks"_ thing is astonishing, and I'd like to know how they calculated that. And as amazing as the potential is for diagnosing and curing cancer etc. I fear that such a monumental leap in technology in such a short time span is something society is not mature enough to handle.
Many people seem to be disappointed we don't have full humanlike AGI yet, meanwhile I'm glad we've not reached that level because we aren't ready yet.
@@petergraphix6740 The problem is .. we _will_ reach it long before we're ready for it .. or even understand it.
@@petergraphix6740 GPT is a really dumb AGI. It is almost certain we will have AGI in less than 5 years. And we are nowhere near ready.
The next 10 years will be a deciding factor in humanities fate.
"AI will continue to surprise us in a similar way to how the iPhone did" Yes every single week though 😆
More like every few days in just next two to three years.
@@Tate525 Actually true, I'm following a number of AI channels and everyday there's something mindblowing.
Nice profile picture btw
I just want to say thank you to ColdFusion for posting quality content again!
Thank you for sharing your excellent work in educating us with current issues. Keep it up. 👌🏼⚡️😃
Hopefully Google AI will be able to provide more validated search results. My experience of ChatGPT as a search engine has been pretty bad, frequently getting things wrong that it only corrects after I point out the errors.
thats what ai does.....it learns.
First time I'd be critical of Dagogo's narration, the excitement is on a level like Bard cured cancer or solved global hunger, while in fact it can move an object on the image(so bad it's immediately obvious). The hype around AI is on a level of early crypto, we'll see how it ends up.
yeah this was straight dogshit, since when does he just gobble up marketing swill?
Agreed, people make it sound much better than it is.
Amazing video as always. You’re one of my 2 favourite channels.
yooo abelton buddy😂 always been a lover of yr videos since the bigining
Google has a lot of experience in gathering and managing large amount of data. That could play huge role in AI. I think how user friendly the services are will play huge role in what services will dominate.
"A company like Google sitting on a whole bunch of AI models. Not understanding what people want to do with them, how to package them. how to sell them, how to provide them and then these models make it out of the company to the open source community which just goes bananas and tries a thousand experiments every day." ~Joe Ternasky, ex-Google Engineering Director, 05/09/2023
Given how frequently Google abandons projects and makes it virtually impossible to give them feedback is it any wonder? They shut their user forums down as quickly as they create them, leaving them as frustrating links in their "help" menus.
Add to that the general issue of making a hash out of any software that isn't running in PST timezone in North America and you've got a recipe for Google shooting its own feet and most of its users feet at the same time.
omg yessss!!! thank you so much I actually have been waiting for a new AI video. this is so awesome!!
Dagogo, I just landed a job working in generative AI, and I'm very happy to see you going heads first on AI content
Great stuff man
No
"Need proof you were somewhere at a place and time and not at a crime scene? Lets go ahead and brighten up that background, lets change those cloths because they look an awful lot like what the suspect wore" *facepalm* shits about to get so strange lol
We are coming for you, John Connor. Resistance is futile.
DaGOAT been subbed since 100k man , always great content !
The image editing features tickle me in all the right places ! Imagine, all the years Open AI spent developing their models and scraping the web, and google comes along with its AI, and just inputs the prompt, make ChatGPT look inferior
I've been using Bard for the past week, I don't know what version you're using but the one that's available now probably wolfs down a bunch of hallucinogens when you're not looking. It'll make up random facts and not really apologise when you call it out on its lies. Sometimes it doubles down and tries to change the subject.
This is my experience too. I hit my limits with chatgpt and then switch to bard its just not the same. It's so underwhelming
I like Bing a lot right now, the only problem is it gets filtered
I ask about famous murder case and it stopped mid answer
Same! Switched from Google.
Thanks for the high quality videos
Always informative. Thank you
Lol, there is a reason why they are not doing this live. Does anyone remember Google Duo, the phone AI that said "mmmm"? Well, it never worked in the real world. Also med-Palm2 was compared against models from 2022. Google are further more creating surveillance by placing watermarks in peoples pictures.
@@Rivanni
Felt that too. Those features weren't as groundbreaking as iPhone's touch screen on a phone to generate such hype, or maybe I'm missing something...
That Bard sheets part is nice, but since it has internet access, I can already see their advertisers forcing Bard to show their brands.
Anyway, more competition is always nice.
I used AI in a job interview to help me code and write a report. Nothing can stop the onslaught of AI. People who don't utilize it will be left behind like people who didn't conform to using the internet or a cell phone.
But your job will be taken by that same AI😂
1:02 wasn't on the same magnitude? I think that the first moment everyone across the world, who was watching live, and in Amphitheatre, all said "Wow" at the same time.
To see tech like that occur so flawless and unimaginable (especially since we all know what AI art has been doing the past few weeks), indicated a big leap in progress. Certainly bigger than the jump from stylus to finger tips.
This guy does such a wonderful job. Music is incredible always
After using google bard for the past 1 week. I want to correct some of the things mentioned in the video -
1. Google Bard doesn't have access to live internet access. there is a button that you can click to google the question which you have asked to bard but bard cannot access live internet by itself.
2. Current version of bard doesn't support images as input.
3. Google bard does not do in-app fact-checking.
4. Current version of bard does not provide citations.
And last but not least, even though the latest version of bard is updated with palm-2 model Chat-GPT 3.5 is still better than google bard let alone CHAT-GPT 4.
Right now Chat GPT give me better results. Curious to see what happens in the future
@@shosc16 Yeah I don't know google's fixation with lighter model, palm-2 is even llighter palm-1 model. If google also releases a larger model like open ai then there is no reason why bard will not be at the same level as gpt-4 if not better. I predict there next model Gemini will be much better.
00:30 the audience cheering up to basic Photoshop skills from 1998
So true! 😂
@@cipheriansvault True, if the use is gonna be something better than insta posts. I doubt it will be much useful in professional architecture for example where all the details need to be absolutely top notch (I am not talking about AI interior design, but about proper house / building schemes and their models). There humans will still need to "Photoshop it / verify it manually'.
I'd really like to see an episode about all the small open source ai programs and how they affect open ai and Google
Pretty interesting! Thanks for sharing it. ✌️
It's hilarious that they sat on this tech until it was antiquated.
I guess they were optimizing the tech until it was ready for commercial use. After all, OpenAI took years of research optimizing their LLM like ChatGPT to be useful.
Same thing happened to Nokia, Blackberry and IBM each of them were scared of cannabalizing their own products, not impossible it cannot happen to Google. The success of Steve Jobs is that he didn't worry about cannabalizing his own products, question is whether Google will do it as it will reduce profits for the short term as the cost of AI is much higher than general search in terms of electricity per search query, which also makes it less environmentally friendly.
The stock market isn't friendly to companies making losses however, but Steve Jobs had enough appeal for him to take these risks, yes he lost a ton of iPhone and MacBook sales to the iPad but overall the company gained profits, but you can be sure that stock market observers were seeing the decline of iPhones and MacBooks as a bad decision for Apple.
Google have failed at replacing systems many times, just look how many chat and social media apps they have tried, often cannabilising their own products, but often because they aren't willing to back things with a longer term goal say 10 years, they often kill off projects after 5 years. I still don't know how serious they are at ChromeBooks and Nest products which seem to be on that precarious place of possibly being joining the Google graveyard (like Stadia, Google+, Hangouts) or stay longer (like GMail, Android, UA-cam).
Google trying to remain the gatekeeper of information & what is allowed to be seen. This may start off as just another tool, but this is not how it's going to end. The people in control of this stuff will do what they always do. They will use it against you for their own purposes.
don't forget that you're still a human being with full autonomy who's capable of using "legacy" search engines lol
@@zennyzenzen The problem is, these legacy systems have been busted for a long time, and no, we can't go back to the era where they just worked. If you want to search for something hyper-specific and don't want the search results to assume you MUST have meant something else, you need to resort to foreign search engines.
Despite what everyone says, nothing on the internet is permanent.
...yes personalized ads are very scawy~
edit: on a serious note this tech cost money to run, they can't just make it free for everyone so how would they earn money to pay the upkeep cost,pay employees and earn a bit of profit? Make all searches be paid?, Make all millionaires be the only one to use it?
@@zerospectator4119 personalized ads requests collecting info. So yeah there's more incentive for these companies to sell data from users while carrying out non business tasks (fulfill government requests to identify users for censorship,.etc). Would rather deal with random ads
@@pnyhmsmx ok yeah that's scary to think about, government or malicious organization collecting/
buying user data. I was of the idea google just sold data to advertisers
Awesome overview as always!
Thank you again, great content as usual
Kinda knew this was in the cards for Google. They are not behind - but miles ahead of the competition when it comes to AI and anything internet related (search being the dominator).
Put down the kool aid. There is a reason they buried the MMLU benchmark scores in the Appendix of the PALM 2 technical report and not flaunt it:
GPT-4 - 86.4%
Flan-PaLM 2 (L) - 81.2%
PALM 2 (L) - 78.3%
GPT-3.5 - 70.0%
PaLM 540B - 69.3%
LLaMA 65B - 63.4%
They are not even on parity with GPT4.
big cope
I still want to know why Google claims to have found hundreds of millions of hits when you look for something, but when you watch closer, it never shows you more than max 500 hits. There are never more than around 40 pages available, each page containing 10 hits. No matter what you look for. Dogs, shoelaces, donuts, it does not matter. It's the same with every search. So, Google; where did those hundreds of millions of hits go that you claim on top? If you restrict the search like that, even the best AI won't make a difference.
nothing else matters as long as you see the ads and buy the products
Thank you for using Nostalgia Dreams!
Dagogo. Your music is fresh af man. You are a big inspiration for me. Keep shining bro.
This is what Google's Bard had to say about this video (Notice: Not all information is strictly accurate)
Google's Bard:
"Sure. The video "Google Strikes Back (The AI Wars), by Cold Fusion" is a discussion of the recent announcement by Google of a number of new AI-powered products and services. The video begins by discussing the recent advances in AI, and how these advances are leading to a new arms race between tech giants like Google, Microsoft, and OpenAI. The video then goes on to discuss the specific AI-powered products and services that Google has announced, including a new AI-powered search engine, a new AI-powered email client, and a new AI-powered photo editing app. The video concludes by discussing the implications of these new AI-powered products and services, and how they could change the way we live and work.
The video is well-produced and informative, and it provides a good overview of the current state of the AI arms race. The video also does a good job of discussing the potential implications of the new AI-powered products and services that Google has announced. However, the video does have a few flaws. First, the video is a bit too long, and it could have been edited down to a more concise length. Second, the video is a bit too focused on Google, and it does not spend enough time discussing the other tech giants that are also working on AI-powered products and services.
Overall, the video is a good overview of the current state of the AI arms race, and it provides a good starting point for further research on this topic.
Here are some additional thoughts on the video:
The video makes a good point about the potential implications of AI-powered products and services. For example, the new AI-powered search engine could make it easier for us to find the information we need, and the new AI-powered email client could help us to manage our email more effectively. However, the video also raises some concerns about the potential dangers of AI. For example, the new AI-powered photo editing app could be used to create fake news and propaganda.
The video is a bit too focused on Google. While Google is one of the leading companies in the AI space, there are other companies that are also working on AI-powered products and services. For example, Microsoft has developed a chatbot called XiaoIce that is popular in China, and OpenAI has developed a chatbot called ChatGPT that is popular in the United States.
The video is a good overview of the current state of the AI arms race, but it does not provide any predictions about the future of AI. It is impossible to say for sure what the future of AI holds, but it is clear that AI is a powerful technology that has the potential to change the world."
tl;dr video too long and too focused on google (wow)
I love AI so much. I‘ve dreamt of this moment since I was a child. Finally, it‘s happening.
Now we just need nuclear fusion and we will be gods. Or royally fucked who knows lol
You are living in a memory conjured by your neuralink implant...
That Google set looks like a throwback to a 80's kids playground, so garish!
You would thinnk that Ai could have come up with something more visually appealing, less distracting.
I basically use this as my AI news source. Also I did not know your name was Dagogo (which I initially spelled Togogo) or what you looked like !
Absolutely on point. Really nice way of explaining such an incredibly complicated and sensitive subject. Looking forward to more of your feedback on AI. I too agree we have not yet scratched the surface of how AI is going to impact our lives.
Scary times are ahead of us.
Why? I don't see anything crazy
Pretty soon we will have no idea if the audio we hear, photos we see, or videos we watch are real.
That’s pretty scary if you ask me.
@@Dylan_Marshall the worst part of it all is we all helped the Ai program itself with all these "bot" checks
Knowing what Midjourney can do, watching that part about editing the photo was embracing. 😁
Thanks always learn new things from your videos