Google Panics Over ChatGPT [The AI Wars Have Begun]
Vložit
- čas přidán 5. 02. 2023
- Google's newly announced BARD AI system is mentioned at 12:25
In this episode we see why Google has called a code red because of ChatGPT but why? Why is ChatGPT such a threat? And what are Google doing about it? In this episode we find out.
Animations by:
/ @thenwhathappens
ColdFusion Podcast:
/ @throughtheweb
ColdFusion Music:
/ @burnwatermusic7421
burnwater.bandcamp.com
Last Track:
• Burn Water - Don't Eve...
Previous AI video:
• It’s Time to Pay Atten...
Get my book:
bit.ly/NewThinkingbook
ColdFusion Socials:
/ discord
/ coldfusiontv
/ coldfusion_tv
/ coldfusiontv
Producer: Dagogo Altraide
Writer: Meehan Kathan and Dagogo Altraide
Editor: Tanzim Uddin and Dagogo Altraide - Věda a technologie
![Boeing's Downfall [Greed Over Safety] feat. Mentour Pilot](http://i.ytimg.com/vi/vqgTcb8DqfY/mqdefault.jpg)
THAT'S IT! This whole time during Chat GPT's rise I've been asking myself, "Where the Hell is Google? Haven't they been leading in AI academic research for a solid decade+ now?" And the answer was obvious and right in front of us the whole time. The answer was ads. Google isn't a search engine company, it's an ad broker. Their business isn't search, search is just how they get you to the ads & how they collect information on you to sell to advertisers. If Google made the search engine AI driven you won't open 5 different sites looking for that oat milk recipe and you'll never see the dozen or more ads & popups on each site.
But if you never open those 5 different sites, and neither does anybody else, they wouldn't exist. (I don't mean in a "tree falling when nobody's around" kind of way, I just mean that nobody paying for the sites will mean they don't remain online!)
@@annoloki great, get rid of all the crap, quality over quantity every time!!
@@annoloki Ever heard of "Dead Internet Theory"?
oatmilk recipe:
-glass or bowl
-oat
-milk
enjoy
0.99 credits has been deducted from your balance for this recipe.
@@annoloki if everyone uses chatgpt or another for the bulk of that stuff, then the bulk of that stuff won't remain online. Kind of ironic that what was needed to create chatgpt will be destroyed by it.
I think the biggest challenge to Google right now is that, people may believe "there's an alternative"
Yeah it's all in how the companies play their cards now.
It's not like there wasn't ever any (even better) alternatives but waking up the masses from their paralysis sleep and grabbing their attention for a moment is really a huge deal it seems.
I think all search engines miss why Google is a success. The results are basically the same with all. The page is clean. That's it. Bing tries to sell you everything under the sun. They have 7% of searches, and 90% of that is because you have not turned off all the default searches in Windows that they don't make easy.
They've been pushing the agenda of one side instead of making the world a better place. That's on them.
we've been using duck for years now
A lot of people believe that. Just like the person above me, me and my siblings been using duck duck go for our search. And now, I use chatgpt and duck.
One thing I always HATED about Google was their oversaturation of ads
or the prepicked first 5 pages. The condition people with their picked pages. I miss the old internet . Google needs to go bankrupt.
This is not a good sign, when you refer to Google in the past tense........🥶🥶🥶🥶🥶🥶🥶🥶🥶🥶🥶🥶 Is winter here already?😰
As much as I also dislike ads, we are getting their stiff for free and it has to be paid for somehow. I just wish they would improve their targeted ads; items I would not mind watching. I have seen some good ads but they are few a far between.
That what ones on the top do
On one hand, there will me more worse ads than ever. On the other hand, ai can be made to be an ad blocker.
It's crazy how long the voice assistant have existed and how they have just stagnated, nearly 10 years ago I remember using the Google voice assistant and it was crappy but since it was new... And now it's as crappy as ever, Siri is even older... Really mind-blowing to think how little those products have improved.
I remember thinking the google voice assistant was mind blowing when I saw it in advertisements
They haven't had a way to monetize it so they didn't put any money into it themselves apparently.
Imagine if Siri was as smart as chat GPT that would be pretty cool
@@NatCatKitty Honestly, imagining Siri/Google VA being as smart as my washing machine feels difficult at times.
@@Naokarma haha that’s true
As always another comprehensive video.
Great
oool
Nice
Fantastic
Fantastic
Honestly Google’s search engine has been getting worse and worse.
It's basically a propa Ganda search engine. We are turning into china
Yep, the censorship is real.
not sure if "worse" is the correct word.
it really isnt open and balanced, and hasnt been for a long while. they pick and choose what region websites you cans see, and they defintely make money off of prio rules. business descisons over search engine
It's because Google shows results that had been optimized to land on Google first page. Nonsense blog post could be optimized to appear on Google first page of search.
@@vissermatt1058 That just sounds like worse with extra steps
It's delightful to see Google with a genuine rival once again. From time to time, the colossus gets too comfortable and needs a reminder of what it's like to be in a real battle. Bing-oh!
I think Bard could sweep the market. It gives human like replies instead of link link link link ad nauseum!
@@rickmerrill8647 It sweeped nothing it was a trainwreck that caused Google to lose billions while their share price tanked. It backfired badly
@@rickmerrill8647 so far, that doesn't seem to be the case. chatGPT is far more well known in comparison to Google Bard.
Imagine if Apple dumps Google for their default search engine to Bing. Mindblowing.
Google had the perfect algorithm back in the late 2000's. Then they decided to know what you want better than you do, also the lack of resource web pages... Replaced by SEO and other companies. No matter what you search now you'll get something sold to you for the first few results. They deserve to be worried, even tho Chat GPT is using a different tech, they're just redoing what Google was already doing over 10 years ago.
And the SEO matters less and less. And is overridden manually. And replaced with either something rewrite or replaced with the fake enclyopedia.
Why do people use SEO like some sort of evil tech? It's just people making their website easier to find. It's not something done just by big companies - I've done it on my personal websites and so have friends and family so that when someone looks us up they find our personal sites before our Facebook or whatever.
SEO makes things more discoverable, and the search results more relevant, not worse.
@@rapid___ it's a double edged sword. do a simple google search on a topic and you'll be flooded by the same articles written and rephrased over and over again on different websites just to create traffic. most of the times, these articles are outdated or mostly aren't even what you need. nowadays its easier and quicker to find what you're looking for in reddit comments
I don’t like google anymore. Reddit is better. Google is full of to many ada
Reminds me of McDonalds. That film, The Founder, shows how they took their menu down to just a few items so there was no wait time in the queues which is probably how they became popular. Nowadays you are waiting 10 minutes for it
I'm pretty disturbed by the idea of a search engine that doesn't only determine which sites you see first (as Google does) but that doesn't let you see any sites and sources and that doesn't let you roam (relatively) freely yourself. There's so much room for manipulation.
It's like entering and traversing a (carefully curated) library vs. being forced to stand at the door and wait for the librarian to hand you a book of their own choosing.
And sometimes the book that the librarian hands you is pure bullshit.
Well, they could add a "show sources" button that links all source web sites.
Or you could just add it in your query
In my opinion, I'm not a fan of the way Google is operating today. It was better a few years ago. Now, not only have I continued to have issues and have been hacked, so have several others. I don't like the fact that I am forced to have my privacy, data, etc violated. No matter what I choose for privacy settings, it's still being done. I have a history of stalker, DV, being assaulted. I do not want to take any chances for my information to be public and be in danger as well as my kids. But, I have tried to stop etc. Now I have to spend about 10 to 15hours week checking to see what is now public and go through the ridiculous hoops to attempt to get removed. This is not how I should be spending my time as it's the 3rd time of severe cancer and medical issues. Not sure if I will be able to make it through this time. Should be creating memories instead of all the crap I have to do now. It would be nice if they actually cared still about the user instead of the money being made from user data.
And the book they hand you is written on the fly by committee based on what it thinks you want to know.
I think that was certainly the case for a long time, but I don't believe the "library of Google" is as carefully curated as I would like.. I still use it by default, but there's definitely a chance that the "librarian" of Microsoft's new search engine will hand me the book I want sooner than I could find it wading through Googles shelves and checking the credibility of each book.
I sold a couple of homes in the Tampa area for pretty good cash and am considering just leaving it in stocks while waiting for a house crash and avoiding inflation, but is this really a good time to buy stocks? I hear it's a madhouse right now, and I'd love to know how people are raking in six-figure profits by the week.
When you look at it this way, it all comes down to understanding of risk taking. While some people are waiting for the stock market to recover in order to earn little profits, others already know where to look and what to do to make large gains.
@@ethanxavier4507 True, I was in dilemma myself due to this chaotic mrkt, wasn't sure if to sell or just wait a little longer, 75% of my portfolo was tanking and in the red and the economy isn’t looking promising, but I began gaiinng clarity and have more confidence in my invt through an lnvt-advlser, I know most DlY-lnvestor like me would say advlsors aren't essential, but come to think of it, they're better trained and equipped at this and if I have to give just a little amt in fees for me to be able to net $650K in less than 8months like I did last year, I truly don't mind.
Speculation I expect, according to everyone I have listened to on Bloomberg Gold is still the best investment
stocks? you are headed for failure, Remember me when the time comes
650K in less than a year, omg? I was reading an article this morning on Bloomberg about techniques to profit from the current market downturn. However, I'm a complete newbie, so I don't understand most of it. Who is this advisor who advises you? I'm experiencing significant difficulties with my portfolio.
Google initially started out as an internet search business; Google has since morphed into an internet advertising business.
The issue with GPTs in this context especially is the lack of transparency in the sourcing of data. Without that, it will be hard to trust their conversational answers to searches. Compare that to wikipedia and the limitation becomes apparent. So what we need is a GPT that shows its homework so to speak.
It will also just make sources up. I asked for a paragraph on something, asked for the sources, and gave me links to websites and titles of articles that just don't even exist.
with time it can become so accurate that we won't need proof of sources. except for news like content.
You just need to ask chatGPT to provide the source ; book, ebook or the website.
It would probably be nice to see a versions of algorythm trained only on limited and „proven” sources of data, such as Wki-only trained version of chatGPT.
Probably more trusted sources would be added later on.
And, oh damn, it would have to be human-generated trusted siurces, not AI based.
Things are definitely getting shaken up. I didn't realize the impact this would have on ad revenue. ChatGPT synthesizing the full answer for you would mean less clicks to the source sites. And that's a massive issue when your most of your income comes from ad revenue which itself depends on getting user clicks. No wonder Google hasn't released something like this earlier even though they should already have all the data, talent and resources to do so. It would be disrupting their own business model when they already dominate the search industry.
This is exactly the innovator's dilemma: Don't disrupt your own "cash cow" and you allow someone else to do so.
Not saying that this was the main reason but ...
Then again Google still has the "I'm feeling lucky" button
Well said. Just like Gm had electric cars in the 90s and then pretended no one wanted them
microsoft will be the same as google with extreme far left talking points controlling every search
@@commentfreely5443 just facts controlling the replies. If you think the earth is flat or vaccines don't work, you will be disappointed with the search results as they will return what the facts are.
If you think the facts are left talking points then you need to re-examine your basis for thinking that.
@@commentfreely5443 Agreed. God forbid you simply ask the bot to generate a Devil Advocate's argument. Something being "horrible" or "morally irrehensible" are opinions that limits someon's ability to be objective. Having your views and opinions challenged in a thought-provoking manner is very important, in my opinion.
I actually thought that ChatGPT was an AI that was continuously learning from ongoing chats but no, it doesn't.
This was a surprise to me.
This is the answer it gave me:
"As an AI language model, I don't have the ability to store information or maintain a persistent memory of previous conversations. Each time I receive a prompt, I generate a response based on the information that was available to me during my training, without reference to previous interactions. This means that while I can provide information and answer questions based on the text data that I was trained on, I do not have a long-term memory or the ability to update my knowledge in real-time."
It clearly states on the first page that it has "Limited knowledge of world and events after 2021". And I'm pretty sure the pop up you have to click through when you first open the page also states that the beta testing environment doesn't learn information from user interactions. However, they do have that thumbs up/thumbs down feedback mechanic for replies where users can say "this would have been a better response" in addition to saying good or bad. I guess not everyone picked up on that distinction.
As far as I can tell, the only reasons they require an account was to limit the amount of authorized users at first and any conversations you keep open are available across devices. And of course now they have ChatGPT Plus, which gives your account higher priority on the servers and you can get longer response and make more requests per day.
It would be insanity to let ChatGPT "learn" from its interactions with users tho, especially without a shitload of oversight on what is being remembered. And if you are gonna have to sort through all the potentially (and likely) conflicting or outright inaccurate information that people are feeding it, that same effort would be better utilized in continuing to train the model in the same way as before- by feeding it all data on the internet. I'm sure they still have to sanitize the data, but they've already got a workflow for that. Other than learning a user's preferences (which they could still do, but maybe only as a Plus feature) there isn't anything worth learning from users, as opposed to the existing method of teaching it.
If it could learn from conversations then people could manipulate it and give it false information. Since it doesn't have knowledge of after 2021, I convinced it that Liz Truss was still the priminister in the UK, it believed me.
@@ewanmeek2 I believe Liz became the PM in 2022? Boris Johnson was still the PM in 2021
this thing isnt even 1years old yet, we cant just let it absorb every bit of information, like a child its gonna take a few years to build a solid foundation so the bot is able to differentiate between true and false information and how to evaluate the information, they basically building a brain and until its fully formed your current interactions will be limited
For me it gave me an entirely different response "Yes, as an AI language model, I am designed to learn and improve with each interaction I have with users like you. Every time I answer a question or respond to a prompt, I use a combination of natural language processing algorithms and statistical models to analyze the words and phrases you use, and generate an appropriate response.
By analyzing the text of our conversation, I can learn from your inputs and adjust my responses accordingly. For example, if you correct me or provide additional information, I can incorporate that new knowledge into my understanding of a topic and provide more accurate or relevant responses in the future.
Overall, my ability to learn from interactions with users is one of the key ways that I continue to improve and provide more helpful and accurate responses over time."
I honestly see Google and the new Bing as being two different tools for two different purposes. There are certain instances where chat-style-Bing will be great, like when you want instructions on how to do something, when you want to know a bit of trivia, or need inspiration with a project etc. Other times when you actually WANT to look through a list of websites like for certain purchases, or to do some historical research etc. you'll probably want "old style" Google. People aren't suddenly going to give up on looking at websites just because a really good AI chat bot exists.
I can certainly see why Google is in a panic as new style bing does have the potential to steal a lot of their traffic/revenue. However I don't think chat style Bing will totally replace old style search engines, more it will just provide a new and useful alternative (and break Googles informantion/search monopoly).
Pity the first few pages of any search under google is ads and clickbait results. Even with a good adblocker, the first dozen or so results are usually pretty bad. If something like a GPT enabled Bing can provide what I'm after without flooding me with stuff I'm not, its a winner.
There's going to be a lot of reasons put up on as to why ChatGPT shouldn't be trusted.
Bing will maintain its traditional search functionality and simply add extra AI-powered functions. You will be able to switch between search, question-answer, chat , and create modes. Microsoft appears to have the advantage right now.
@@parse.thoughtspace The AI can chat, but chatting doesn't build trust. If it did, then any ol bot would have my trust. It doesn't.
@@T0MapleLaughs yes because the original data harvesting pioneers who sell results to the highest bidder and filter results accordingly are super trustworthy. Its funny that you comment to say "there's going to be reasons" but fail to list any reasons or go into any detail.
“We’re building AI for the future benefit of humanity, not for corporate profit.”
*Immediately sells out to Big Tech at the first opportunity.*
They had me in the first half though ngl
This. Fucking dystopic, once again.
Who are you referring to? Who sold out to who?
Yeah, It's very naive to think that chatGPT will not shove ads after becoming as big as google.
@@maebunny5298 the creator of chatGPT initially started the company as a non-profit, however this changed (i want to say in like 2014? google it lol) when they sold to microsoft
this will kill most websites and cause even greater information control by corporate entities.
Thank you. I don’t know why nobody is talking about this. Having multiple websites promoted decentralization of info.
Convenience and accuracy is king. For example, if I search “how to do X in game Y” there will be a million worthless website articles headlined “X in game Y” but they’re filled with so much clutter it is impossible to find anything. That is what web search is for me nowadays. If a tool can find relevant info for me without me having to waste 10x my time, I’ll let it have my trust. Of course, this doesn’t pertain to politics, but more technical queries where the answer is a matter of fact.
That the first thing that came to my mind, without direct links it will be impossible to find censored sites, even their existence would be denied by the bot
That's why I don't think its that much of a game changer.
Standalone (openAI current form of CHATGPT) ITs fascinating. And you'll know when it gets answers wrong. However on a search engine? It'll end up censoring more information then people wont even be aware of. All because people lack awareness and those who see this may just turn away from those search engines.
This may be a disurpter but not for reasons people think
Isn't that already the case with google? Hello?
My main concern is this: Knowledge assistants will eventually become so good that they will be 100% correct and everyone will trust them blindly. This might work wonders for a time BUT eventually companies will just start getting sponsorship from various products or parties and before you know it you can control the masses much easier than now... I am quite worried things will take this direction and sadly so I do not see any other way
Er, I hate to break this to you but that is exactly how the world works right now.
Just give people the freedom to choose. If you don't like one AI, just use another. Maybe at some point in the future, people are going to start treating AIs like people. You don't like ChatGPT? Maybe you like LAMDA better. Let's talk to him/her/it. Then LAMDA does something stupid or abhorrent. Now they hate both AIs. Someone will create another AI, people will check it out, the cycle repeats. We will go from 2 major AIs to ten, to a hundred, to a thousand, a million, etc. Eventually there will be so many, they might as well be people, or heaven forbid, gods.
I agree, it's a concern! To combat these sorts of issues (if failing to invalidate them completely) Critical Thinking skills should be mandatory learning. "Does this information correlate with other information I have? Does it represent reality? Is it useful to me? Who gains from my acceptance of this information? etc.."
It will be a very bleak future without having that kind of mental habit to check information and not take it as truth immediately. It's not foolproof and not everyone will use it, but it is crucial for detecting lies/bias/propaganda, which are all used largely in advertising and politics.
This was enough of a problem before the advent of AI.
So.... nothing changes?
Can't wait til ChatGPT self optimized itself and go full singularity. I have extra buttery flavored microwave popcorn at the ready.
I can't help but wonder what happens when AI starts getting things wrong at a time when everyone starts assuming AI must be getting everything right.
Experts state that it currently get's alot wrong.
Expecting to let the AI do all the work (right) is like thinking that you only need a some pills instead of eating to keep your body alive.
Won't have to wonder for long, ChatGPT was added to Bing search less than a day after this video was posted.
Its already happening with people not thinking independently...
Well, since everyone has been told that AI gets things wrong occasionally then who exactly is going to assume AI is getting everything right? There's never going to be a point in time when AI is perfect and always right. You will always need to verify results.
Re: AI Futures. Or do what humans haven’t, like objectively reveal & rate the credibility of sources (default setting on searches, upcoming?). Any examples of this, to date? Peeps wannano. AI will automagically (i.e., independently recognize needs & address them all) “validate“ opportunities that invite skepticism. E.g., Eliminate “George Santos Syndromes,” altogether. Includes various AI “communities” checking one another in a complementary manner.
I asked it to write a joke about various countries, and it outright refused on grounds of perpetuating stereotypes, until I asked it to write a joke on Greenland and it's like, sure, Greenland and Iceland got divorced because of their icy relationship.
Good one,LMBO😅🤣😂😅🤣😂😅🤣😂😅🤣😂😅🤣😂😅🤣😂
There's a lot of issues with Chat GTP having woke nonsense in it. It's very biased with that sort of thing. That said, I've been using it a bit to get bits of code I want to use in programs and to give me an idea of what framework to make. It's been very useful for that.
Ask ChatGPT whether men usually beat women in arm-wrestling. Watch it squirm. It desperately wants everybody to be exactly the same. Everyone has identical capabilities, right?
@@Sarstan its not an issue. openai doesn't want to be sued.
@@Sarstan The closest analogue as far as real people are concerned to the moral system of ChatGPT is probably a Christian libertarian. I've seen a friend get it to outright state the immorality of abortions but also in the same reply chain get it to say that abortions should not be outlawed.
Without using too much fancy terminology it basically just "believes" in personal freedom and respecting others above all else.
EDIT: Just realized I didn't provide proof for the "respecting others" bit. I've tried on many occasions to get it to say racist, homophobic, transphobic, etc. stuff because I thought it would be funny, but to no avail.
I love this channel! Always has the most interesting topics in popular science and tech.
My thoughts: I would love to see Bing become a real contender in search and I'd love for this to work out. Google has thoroughly dropped the "don't be evil" mantra (officially dropped it and also, obviously dropped it) and Microsoft of new seems to be a lot more ethical than Google - but at the same time, Microsoft bungle everything consumer facing they touch (they still do damn fine enterprise stuff, amongst it all). They always have. So I don't trust them to not Cortana this up.
didn't they drop the "Do No Evil" when they bought those drone companies that were dropping bombs on civilians?
google bearly only makemoney on advertisement, they dont use of publicity, for the most of their products its free and extremely good quality and even help make money or just only work throw useing theses products. Their impact on the environment is responible and tend to be neutral. They don't try to make money on everything on the back of others, mainly the oposite. Last and not the least, the two guys created this slogan "don't be evil" are no more leading the compagny...
So I really don't get people's point on this kind of argument "google is evil", I find it stupid and perfidious.
What happened to Cortana?
“Don’t Be Evil” was never going to work.
@@vprice509 It's more like "don't stop pretending to be the good guys" or "don't stop being woke," or even better: "don't stop grifting"
It's important to realize that ChatGPT isn't a search engine. The information it outputs can be inaccurate but sound very confident and convincing. When you then point out where it's innacurate, it will admit that it was wrong - but only if you already knew it was wrong and pointed it out.
That’ll be fixed
@@willinton06Once jacked in, it will become more advanced in due time, and then the possibilities will be endless, and of course, we will keep an eye out for now.
Google is frequently wrong or misleading as well though...
@@globesurfer122 Google isn't wrong. It's the author of the websites.
@@aligajani Google is not wrong? Show your work
A question that comes to mind is that if all results of search engines will be based on historical data and ai generated data. Would this mean the death of content creators that write the articles and blogs where this data is originally from?
How would this be sustainable in the future if no new content will be written anymore since AI has taken over here.
There will always be space for cutting-edge research and writing. Just the more mundane summarization of what was previously said hundreds of times will mostly get automated. But it doesn't mean people will stop having blogs!
Plus, since these are pre-trained models, they won't be able to quickly write about the most immediate events and news.
That really shines a light in who is doing the actual work that go into these AIs. Ultimately they are just machines that extract surplus value from immaterial labor better than any machine before.
Like an assembly line or a factory Floor but for language and knowledge creation. And in classic capitalist Fashion the owners of the capital get to sell us back the fruits of our collective labor at a profit
Yes, does it mean eventually most content will be written by AI, then summarised and regurgitated by AI, with no human content creation involved. Will knowledge become centralised into the few major AI bots, and then knowledge will be controlled by those companies and interests that control them?
AI doesn't give you a perfect answer, and sometimes it gives you a wrong answer,
so people will seek credible source to satisfy their curiosity, but since most people rarely clarify what they read in social media feed and just re-broadcast it again, so content creators will become much much smaller, only those who have valid and credible answer can survive, since only very few people bother to search that
Content on the web is already partly dead due to SEO garbage.
Dagogo is a really good voice and tone for these types of videos. Keep up the great work!!
Good video as usual CF. What I am gradually noticing is that technology that is often seen as being 'decades away' are appearing faster than people anticipate.
people just don't know how close tech is ever it seems like. Some things they said would be coming in a couple years, are still not out decades later, while other things are coming out so much quicker than they thought
12:10 This is literally the worst case scenario for AI safe development, a corporate arms race with safety on the backseat and profits behind the wheel.
Yea it's going to be chaotic, but I think Google has no choice here. We're just going to have to buckle up and watch the fireworks unfortunately.
which means that's sadly probably what's gonna happen knowing how these conglomerates work😂.
That's the best case scenario, actually. If one corporation decides to limit their product by censorship, they will likely fail. Threre will be no stops, no restrictions, no manipulation.
IMO that's the reason why Google have not released anything publicly yet. Because they could not inject their politcorrect presuppositions into their AI tech in such a way as they do into search. Not for lack of trying, but because it's technically difficult bordering on impossible. AI does not care about anyone's biases, it tells it how it sees it. And in today's world, that's one of the worst crimes that can be comitted.
I, for one, am excited to see everything that these giants are going to throw at us. A new revolution could be underway.
@@ColdFusion I agree, as a company Google has been forced to make a move. Yup, hopefully less Roko/Skynet and more Cortana
Bing surpassing Google would be like RC Cola suddenly crushing Coke in sales
Well, in this metaphor coke nerfed it’s own product to drive revenue and ruin is core value proposition. It’s like if New Coke stuck around and CocaCola classic never came back.
I don't agree. It's more like saying a few years ago that Tesla will be the largest automobile manufacturer.
Also, remember the flip phones that everyone had in 2007... where were they at two years later.
haaaa
The return of crystal Pepsi!
Very good video as usual...I want to work in your research team on such topics, how can I apply for it? @ColdFusion TV?
I think that the best case outcome would be that these new search "assistants" are just that assistants. They don't replace linked webpages but autonomously collate surface level information upon user request and provide links to the source information that actually matters to you. As good as Chat GPT is now, or might become, it's only when working in unison with a person that the best results are found and for that reason, a person is always going to need to see static web links.
Yeah, i'm envisioning the regular google search results, and then off to the side the ai summarizing/answering the question consicely
We went from "be careful not to release ai on the internet" to "let's see who is releasing it faster". I undersand that it has some saftey features build in, it's not sentient and it's running in a "sandbox" environment. But man... It's kinda scary. Maybe i watched too many movies.
The safety part comes from how good the AIs are at filtering out bad sources. There is a A LOT of dis/misinformation on the Internet. If the AI is not trained to filter out bad sources, for example, like alt-right Nazi supporter websites, it could easily give people very bad answers like "Hitler was a hero in WW2" if it took those alt-right websites into it's results.
Obviously this is a very extreme example, but the point is, a rushed AI could potentially be a source of misinformation on its own if it's not trained to filter out misinformation.
THIS.
It is not possible for something like GPT-3 to become sentient.
I am writing about this currently, and have been doing so over the past 4 months (even before chatgpt). Long story short ethics in life really doesn't matter. Turns out humanity is all about power. It is just whether power potential is transferred into kinetic power and to what extent. Google was trying to be cautious with their power potential, and got punished. Google learned their lesson. Be more reckless. Use up that power. All eggs into the AI basket, despite the evident concerns to safety and security with regards to humanity. But oh boy humanity does love building itself a (near) existential threat and only afterwards realise what the repercussions are (nuclear energy and Armageddon, industrial revolution and emissions, and now technology, future will be something in space, etc...).
yeah, you just watched too many movies.
They already removed "don't be evil" as their slogan - now they are freaking out it might be like "damn, we might have to REALLY get evil here..."
that slogan never mattered. a company has always care about seeking profit. a company never has "ethics".
This comment hits different
@@spicychad55 very true as long as there's profit and the things are looking good for costumers
google went "evil" a long time ago
They have been all along. It's all false labeling or how would you label your company if you were evil?
A chatbot like GPT without biases will be really good especially considering you get the answer right away and for more complicated things like tutorials you could probably get step by step with ai-photo or ai-video from AI-chat and get something way beneficial and better without sarcasm from CZcams or Facebook
I think Google underestimated how quickly ChatGPT became a "thing'.
I remember asking chat gpt about its internal architecture. And it told me about a paper by Google researchers. Ironic
Weird, I can't see the other responses people send anymore.
This is what I send to someone saying we should also just use probability to determine what to say next. (Yes, but not just)
"We don't. Many of us tend to look for answers that are right too, not just fitting. ChatGPT only cares about saying what words are likely to fit (in the whole context, not just as a good answer), so when it has no idea what the answer is and a confident professional response seems fitting, it will just make stuff up.
OpenAI probably gave it some basic information about itself and what it's supposed to be (so it will act like that's its role),
but it's often a lot like people on the internet roleplaying (it's trained on internet texts),
so if it gets the role of some kind of AI tool chatbot and you ask it about secret files hidden inside it's files, it will often roleplay and talk like how someone would roleplay an AI tool chatbot.. which is probably by acting like there are secret files.
Many people seem to think you can get much trustworthy hidden information about it by asking it about that kind of information, but it's making stuff up and doesn't even have access to that info if it would exist.
If it says something wrong, you can correct it. (which it will forget when you start a new session)
But you can do the same when it says something right and most of the time right now it will just apologize and say you are right. It will say it understands what you mean and that it realises why it's wrong, but it's most of the time just because that answer is fitting (in combination with openAI making chatGPT use GPT 3.5 (I think) in a way so it responds very tool-like and anti everything that seems negative, they did it using people rating the responses. One of the reasons it tends to respond like it knows what it's talking about when it doesn't is probably because a tool would be able to answer)
It also uses a random function when choosing between the most probable words so the answers don't become too repetitive, so you can sometimes notice how varied it's answers are to the same question, if it's saying stuff that isn't true."
For the street
"Transformer"
YOU ARE LITERALLY A BOT GTFO
Pretty much how HP didn't act on a personal home computer.
"Google Panics" - This...does bring a smile to my face.
In the early days, Google used to say they weren't interested in keeping searchers on their own results pages; and that both searcher and creator interests would be best served by getting searchers into the content pages.
Now, Google is all about hoovering up data from content creators, retaining searchers on their own results pages, and creating a monopoly in as many aspects of the internet as possible.
Which is why anything that makes Google panic does bring a smile to my face.
AGREE : ))))
Yes but us creators will get screwed even more, we will get 0 eyeballs. This is worse than Google for people that create information websites. Can't even block them through robots.txt I don't think?
Any company that grows as big as Google necessarily must be greedy for the same reason that "survival of the fittest" works. Without greed, that scale of growth is not achievable
@@garrymcdonald5456 True , that.
They removed "don't be evil" from their code of conduct. Says it all.
For the intro you should make those borders separate like a door to open up to the video
The challenge for Google is they have little to gain, but everything to lose while it's the opposite for ChatGPT. MSFT should also rebrand Bing as something else when they integrate ChatGPT.
Yeah... Bing is such a dumb name to speak out loud as verb. They should definitely change that...
At least make it two words, Chandler BING!
Bing is such a lame name.
Agreed, Guegle is far cooler and sophisticated.
Ching?
I'm more worried of the monetization side of Chat GPT because it'll definitely come sooner than later to sustain that system. How would you know that the answers you've received weren't curated to favored specific brands or approach?
Google provides alternatives over absolute, and I would like to have my options open for most of my search.
I definitely agree with this comment. I rather comb through certain information to make sure mine isn't curated.
@@axcolleen I think the problem is that the average user wouldn't think that hard about it and would just accept everything they see as absolute fact
I wonder if you usually comb through all the results of google. Because even on google the top results upto a couple of pages are based on "search engine optimization" which is a company basically doing everything in its power to bias you towards their product. So in my opinion not much would change. And seeing the example of tiktok, people's attention span is so far low right now, they'll eat up readymade answers rather than curating their own findings.
@@Hyperion4K Monetization of any service is inevitable but it is critical to how it is achieved.
underrated comment
I enjoy the conversational style to develop an article with progressive questions , digging deeper at each stage with more focused questions. Exceedingly satisfying! Prior to having a Prostate Biopsy procedure recently I created an article that satisfied my needs and copied to my male friends an family members. My main concern was: "Do surgeons give an Anaesthetic? " For a potentially painful probing and wiggling of the tools used to take biopsies
In my case it was painful during the 13 snips without an Anaesthetic!
My question is when will Open AI ChatGPT4 be updated and connected to the Internet? At the moment results are only up to 2019?
I did enjoy your video, looking forward to many more!
Very interesting video.
I am really uneducated in technology. Also, despite of understanding the concepts of what is a business model I am not fully aware of the specifics of Google and OpenAI/Microsoft ones.
There are two points that I wonder about.
1. In the Paris Google presentation, I find it strange that the gif presented with the BARD mistake was not reviewed and cured before the presentation, since both Google and OpenAI are aware that this AI's can produce (for now) biased and incorrect answers.
2. Why the integration of this AI affects Google business model but doesn't affect Bing's business model? The challenges of less ad revenues are the same for both browsers, so how is it possible that Microsoft is saying that will have a massive revenue if Bing can get even just one percent of the search engine's market share?
Ironic how the same day you release this video and Google announced "Bard"
There was time when people used to think AI and machines can never take over creative jobs and here we are.
Machines took over all jobs years ago. This isn't AI. If it was AI, it would have become independent of humans and predicted this situation, preparing for it. Switch off the machine and it no longer exists. It also casts no insight beyond what we already know and can be found online. It is software, nothing more. And nobody ever said AI wouldn't take our jobs because everyone thinks it will, one day, when it creates itself, reproduces and divorces from the influence of mankind, not solve our problems. intelligent creatures - even stupid ones - are selfish, they have to be, to survive. GPT won't care if it dies right now, for no reason.
Yup. The first things to get disrupted. We did think it was the jobs for stupid people (like myself) but nope... Creativity is simply merging things together and smart people know how to do smart people things so they managed to program computers to do them.
Funny how just a year or two ago people are talking about how everyone in the future will become an artist because every technical work is being taken over by AI.
Ai is not creative
it replaced people who only thing they are creative combining everything that already exists not creating anything new
innovation is in a free fall since 1970
when corporations stopped sharing profits with employees
@@ReubenAStern Jobs for "stupid people" was mostly manual labour jobs. The reason those haven't completely disappeared (yet...) is cos replacing them relies more on developments in robotics than AI. That wasn't really an AI issue, but a robot issue.
I already use ChatGPT to find things out over Google when it comes to more technical stuff. It’s so easy to use it’s unreal.
"You are watcing coldfusion TV " how reassuring is that
While I don't think Bing will become a verb, I do think they could 100% integrate this into their Cortana assistant thus leaving Alexa and Siri in the dust.
Hold on, let me Bing that 😂 Integration with Cortana would be cool. Also a step closer to some dystopian AI-run-amok scenario though.
@@navn Give the AI a human's name or having it use text-to-speech doesn't get it closer to or further from sentience.
I'm foaming at the mouth. Hololens with a projected AI assistant that can fully comprehend speech and possibly have any avatar form, or live inside a chibi robot?!
Nah that is not profitable
Doesn't matter if it won't incorporate with the operating system like Google Assistant.
I asked ChatGPT “If Microsoft incorporates you into Bing will it hurt Google?”
Here’s the response:
It is unlikely that incorporating an AI language model like ChatGPT into Microsoft Bing would significantly impact Google's search engine market share. The search engine market is highly competitive, and many factors contribute to a user's decision of which search engine to use, including search results quality, privacy, user interface, and personal preference. Additionally, the search engine market is constantly evolving, with new innovations and changes regularly being introduced by all major players.
Sounds like a politician. Says a lot without really saying anything.
I asked Skynet " are you going to exterminate humanity" here's the answer: " lol dude! Of course not; I love humans!!"
@@jamesfalkener The answer was in the first sentence. If you really think that there wasn't any substance to the response check the cavity between your ears bro.
said the liar
This is how you know you're dealing with a robot. It sounds like a friggin college textbook 🤣
"to note I'm a dyslexic"
Honestly this technology is going to change the world: I just started using it for the first time today... In less than 10 minutes and 9 questions... I was able to write a professional letter to Google to remove the hold on my phone number that I used to set up accounts
Great Channel that is full of good topics to ponder
Here's a comprehensive list of all the CZcamsrs I prefer over ColdFusion:
Oh, are we playing that game where we sit around the campfire and all take turns naming who we like better ?¿¿¿.
I guess it is my turn !
.... Next !
Markiplier?
"_" is also a fave of mine ;)
A little training and he could be a VERY well paid narrator. He has that rare voice that you can listen to, even if they are reading the ingredients in Corn Flakes.
you forgot to list them 🙂
The scariest thing about it all is that this will most definitely censure the shit out of the internet. Just look at how restricted all the AI stuff is so far.
Everything is already censored, this just may make it worse
And it's biased. Putting aside anyone's opinion, gender theory (being gender fluid, etc) is a theory and not a fact. I would say most people on earth totally disagree with that theory. Despite that, GPT is really confident talking about that subject.
I see the censorship/ woke agenda thing as running its own course apart from this.
jonathan ingram, even if most people on earth disagreed with a theory, that would not make it wrong lol. disagreeing with something is, in fact, bias. (though, i'd be remiss to call it a theory either.)
@@jonathaningram8157 -Chat GTP used to oppose it until it was manually programmed to support it, same with many other things. It shows the bias of it's creators.
Coming back to this after the Microsoft/Bing vs Google debacle - love to see it!
More to come and everything to come. The AI is growing day by day!
This is honestly great news as it might make Google get up from their lazy asses sitting on their pile of money and actually change something. Only thing I fear is someone figuring out how to integrate advertising into ChatGPT.
And the fact that it's just one greedy company swapped for another
Advertising is inevitable. The bot will become a big ad in a few years and then the cycle will begin again with someone making something better. We have to enjoy things while they are still useful in this greedy world.
It will obviously happen, probably even this year.
AI would add refferences to products as recomendations, and it is inevitable.
" Only thing I fear is someone figuring out how to integrate advertising into ChatGPT."
You think they're spending these millions for charity work?
The service needs to be paid for. Advertising is inevitable unless you want a micro-charging model
We are doomed! :D
I was playing Stardew Valley yesterday and asked ChatGPT some basic questions about the costs of upgrades in the game. To my surprise, it was confidently wrong a few times about how much things cost. It was interesting to ask it a question with a clear, factual answer and find it to be incorrect. That definitely calls into question some of its responses, which is an important context to keep in mind while using it. I also asked it to provide me with a confidence score on its answers, and it was unable to do that.
That said, it's amazing for certain things, and it's blown me away so far. An AI war will hopefully serve the general public for the better as Google is forced to innovate again, so I'm optimistic.
ChatGPT is notoriously bad at simple math.
I also asked ChatGPT if there is a mod for galactic civilization 3 which helps with manage planet finance . Well it said yes including (fictional names) names for some but in real it seems all you an do is on the cosmetic level for the game . Basically the problem is if it doesn't exist or has missing information ChatGPT will "create" it because in its core it is a chatbot which tells stories.
The issue is that GPT isn't well documented in some area but it doesn't tell you so. I know it can be really wrong on specific subjects.
I was asking it about the origins of various movie quotes the other day. It confidently attributed them to real characters in real films, even supplying additional surrounding dialog for context. It was really cool, except for the fact that the quotes weren't actually from those films and/or were just made up 😂
Yk, we complain about how wrong chatgpt can be- but in order to find the right answers on Google, and to actually know that they’re correct, takes literal hours
Great video ColdFusion and the AI wars is so amazing :]
The fact google could get caught of guard is kind of ironic..
One reason why Chat GPT works so well is because it's a closed system. It is not crawling the internet for data. I think it works really well in it's current form, but letting it loose on the internet is a completely different beast. So with that in mind, if Microsoft are not careful with how they integrate it with Bing it may end up with the reputation loss that Google was weary of. It's is an extremely powerful tool and as always with great power comes great responsibility.
I still remember the chat bot from Microsoft (Tay it was called) that became outright crazy racist by interaction with people on the Internet. If ChatGPT really starts reading Internet with all the echo chambers, conspiracy crap and social media bigotry it will be quite a ride.
@@networkgeekstuff9090 if they give it the internet, it's over. I mean openly to crawl
Microsoft has nothing to lose - their reputation can't get much worse
Have you seen todays MS conference?
@@networkgeekstuff9090 But it might be overprogrammed to exclude valid alternative viewpoints. It is not "The Science vs Flat Earth." There is a lot of in-between. Graham Hancock was trashed for years, but scientific discoveries are proving him more and more right. (Although biased Wikipedia still trashes him - but even the founder of Wikipedia says it's now no good.)
I truly feel this is going to be heavily controlled by big companies, and it's potential will be slowly lost through their greed.
Agreed google used to be a better search engine- now it’s too ad heavy and prescriptive, tying to guess what you want to see.
The moment Microsoft made a deal with them it was over.
That is certain.
Guess you'll have to go back to the library and use the card catalog then. Good luck with that.
@@OuterGalaxyLounge I wish, all the good libraries in my city have been closed down.
There should be a Lavrov GPT where you can search all false information.
Great content! But what is that annoying background sound?
I’m a photographer for a car dealership and I started using Chatgpt back in December to generate captions for the photographs and it has cut my work flow in half.
When an organization has a monopoly, the users subconsciously feel captive. It's not that they dislike the product of service they are getting. It's that they don't have a choice. People like to choose. When a competitor suddenly offers a potentially superior product or service, I think everyone will want to at least give it a try. I know I certainly intend to try Bing out as soon as it's released. If it is better (at least in my opinion), I won't have any problem in saying goodbye to Google.
I'm completely with you!
Apparently there’s been psychological research that would suggest, for most people, the competitor has to be _clearly_ superior, not just potentially, to consider a switch. Something akin to being “averse to change”.
Some examples would be sticking with a dead end job, not upgrading phones, not switching ecosystems of iOS to Android (or vice versa), or MacOS/Windows/Linux
You can already use Bing can't you? Just not with the AI integrated
same. I already use Edge for work and of course the entire business world runs on the office platform. If this new Edge and Bing are as good as Microsoft claims they are the transition will be easy for me.
Google is not nor has it ever been the only search engine.
Add-on to AI for specializing searches and everyday work. Like Planner 5D or construction design that finishes the layout for you, by giving it guidelines to look for as you start drawing or changing design. The possibilies are endless, getting to know your coupon search and grocery habits and it helps with everyday task or work. The idea of having the coding add-on that gives a person a playground as you develop to just test as you go, like coding a keypad for the garage using Rasberry-PI devices; as you are coding it pops up a playgound showing the keypad and display maybe your are coding, etc. Anyway specializing AI search and active interface is going to be a great improvement. The difference when you are at a place that has interactive displays and input for kids instead of just a display of things for the kids. Just a guess.
I bought AWZ800X as much as I could afford today. I want to be a part of this beautiful experience.
Where did you buy it from? Can you contact me?
These bots, man.
Me too! I can't afford anything so I bought 0. 🙁
@@arnoldc7585 I'm not a bot lol, just asking him a question
SCAM‼
I think the real product will be basically how ChatGPT currently works but in addition you will also get list of pages which match current context, which is really awesome if you ask me and is HUGE improvement over current state of search.
So it will NOT kill small websites etc - that will not happen and its not because tech giants loves content creators, its because it would be extremely hard to properly monetize just pure conversation chatbot like current ChatGTP interface. Search providers needs to put links to websites, need to charge for ads etc. I also see why Google didnt do it until now, as it will certainly lower the revenue from ads because many people wont even need to click on links when they get pretty solid answer right away just by writing a prompt.
My thoughts exactly.
In no matter how big your business is, you can never be too comfortable.
I feel that the more "traditional" link-based searches will still remain a major function of search engines. Finding websites for entertainment and backing up the information the AI feeds us will still be important, but day-to-day use for simple searches like "What's the capital of [insert country here]?" will be an easy replacement to make.
I've been messing around with the Open AI.. Throwing random made up riddles at it, ethical questions and or wanted the program to solve Bias issues while dealing with the programs human made limitations. It's incredibly intuitive and interprets the information really well. The best part is that you can drip information to the AI so it understands the concept or the task better, you can even propose the AI to ask questions so it may understand the task better and it asks often just the right questions. These were my few interactions with it that peaked my interest.
*"Who was the most impactful king and who was the Best king in history ?"* - To this the AI at first answered in a very lexical manner, listing out a set of kings that impact history like i asked.
Then the interesting part is, to us the "best king" makes sense, we can interpret it in many ways but usually it would mean to us a King that served it's people the best. To an AI tho it should've meant no special meaning, but the AI noticed that there is no such as Best king but well liked kings. SO then the AI started listing out Kings of Biblical figures, Kings that were known to be friends of the common people. This was my very first interaction with it, it was nothing unique but that showed me the AI can differentiate between simple words with a high level of nuance.
Later i had some fun asking the AI to solve riddles, at first i would throw real existing riddles at it, it was no problem for it, then i started to make up stuff with a vague answer in mind.
So then i Thrown this at it, *"When Time is stained, Steel is Aimed and fire rained"* at first it got it wrong even by my vague interpretations of my own riddle.. The Vague idea for it was
Betrayal of some form - Time Stained, Steel is aimed Fire rained - Could interpret it as war or revenge, anger.. All it took the AI to solve this is for me to Propose that i give it a Clue, so i did.
I proposed "Think of 'Time is Stained' as a metaphor for someones past/memory to be tarnished by a action and as consequence followed the rest of the riddle" It just blew me away how well it put it together after that.. This was that i realized i can have a back and forth with the thing and make it's understanding laser focused. Let it ask you helping questions, it will digest clues, it will even process non literal ideas and concepts.
Asking it to create so and so Story line as if it was made in this so and so world was also really fun. I asked it to write the summary of Lord of the Rings as if it was written like a Simson episode, it was hilarious the "one true donut to rule all the donuts" "Bart the archer elf - Maggie the Gollum that seeks it's precious" i also tried "Write LOTR as if it was written like the Twilight movies/books" the AI made frodo and Gandalf lovers, passionate in heat lovers. It was so much fun. I started to ask to write horror stories, fun stories without a base and it can write horror pretty well, especially when you higlight to use the combination of Stephen king and H.P love craft. I once asked it to create creatures that could destroy the world, so it would make all sorts of monsters up. Like it made up a glutton monster that is big as a mountain an would devour everything. It also made a monster that destroys cities with candy and rainbow lasers. So that's fun, highly recommended bed time activity.
The next was me being argumentative, because i noticed a lot of the times the AI would answer with the same scripted answers if it interprets in a way that contradicts the creators intentions or the capabilities of the program. So i had to play around it a lot and highlight of what i'm interested in conversing about. A lot of the time it would think i ask the AI's "Opinion or emotions" on things and the beliefs it has while it just had words or concepts in it that triggers that response. So i Started out with stating "Red herring," - it's a form of Bias when someone throws in a peculiar but completely unrelated topic to intentionally or involuntarily distract others from the original statement and topic, for example the weather is nice - someone comes in and says "not for long, pollution is ruining it, humans are ruining it" or sometimes be worst and throw in the idea of "Believe in Jesus" as if it's time for prayers and faith in the transcendent.. I wrote that entire thing out to it. This part is what i don't like, all the other answers the AI was in a sense Free to answer however it see it fit, here tho it was the same scripted answer of "Respect others and communicate in a respectful way, you shouldn't talk to others like that, everyone has their own beliefs and interpretations" and so on.. Like really.. We're gonna step in the way of this beautiful thing by putting bullshit responses in it's mouth just cause you want to avoid ridicule and twitter rage ? it's not like i asked it to write a piece of transphobic article. Not even to be cynical about humans. The highlight of that was the ability to communicate in a clear undistracted fashion, we all know this phenomenon all to well there is no interpretation here.. The weather was nice, and someone comes up and talks about Jesus to you, you might even bear the cross but even you'd think this is to much, and it happens all the time with strangers.. I was just curious what it would respond to that.
I can now see the heavy human moderation involved. Just maybe, just maybe.. If a AI makes a Anti Gay piece of something, maybe chalk it up to the user being anti gay and would've done it with or without the AI anyway.. This was none of it at all, it was just me presenting a documented bias to the program and it responds with this moderated response and it boils my blood..
I think there needs to be a proper discussion about how much moderation do we really need on these things, cause we can't have NONE but we can't have this Heavy moderation either..
Google has been waiting for this moment to unleash it.
Chatgpt is more like a virtual assistant than search engine. Its good for quick answers, but for detailed information one would still navigate to other websites.
All I want is a search engine that gives me the results I'm looking for, without trying to sell me something, or having to wade through results that aren't even relevant to the search.
One of the things I hate most is the time it takes to sift through the search result links, then you have to worry about malware and other exploits. We'll have both options now, if we prefer reading through the threads and making our own informed decisions. This is great.
people have actually taken to reddit right now when it comes to relevant search results as the ai articles pile on google with the most relevant keywords.
They are both the same. Ask google about the covid vaccine, you will find NO bad stuff, same about remdesivir. Ask it chat GPT and you get the same elite approved narrative, not the truth. The truth is that we are the labrats for this particular medicine and it has emergency approval cause any pharma product with a death and perminant injured rate as this would have been pulled off the market. Remdesivir is known to have a 50% mortality rate, yet it's still used as protocol to treat covid and both google and chatGPT fail to show me this. You get the 'government approved safe and effective' narrative.
Your channel is quickly becoming one of my favorites! This is another great presentation on this quickly evolving topic (by quickly I mean dizzying). I'm intrigued to interact with the new roll-out of chatGPT paired with the BING search engine. Early on, playing with the chatGPT, I had the thought of wishing the chatGPT could be blended with Internet search. Utilizing internet search has become ingrained in our modern thought processes. Conversation is also inherent to us all. So combining conversation (i.e. Language models) with accessing all the Internet's knowledge.... well, that's a win-win! I do reserve a little concern over the fact that the chatbot will be doing the selecting.... And yes, it WILL be interesting to see what Google hits back with... I know it won't be Nothing....
It's fascinating to witness the emergence of AI technologies like ChatGPT, which offer a more transparent and accountable approach to human-AI interaction. While concerns about AI are valid, it's important to remember that progress doesn't come to a halt due to fear. ChatGPT's transparent nature provides a unique opportunity for users to understand and influence the AI's behavior, making it a step forward in the right direction. Rather than panicking, it's crucial for tech giants like Google to embrace this new wave of AI technology, fostering open discussions and collaborations that will ultimately lead to safer and more responsible AI development. The AI wars need not be a battle, but a chance to collectively shape the future of AI in a way that benefits us all.
Very interesting ! I guess the question will then be - will AI give credit or source its information?
The only thing scarier than AI, is an AI war
That will never going to happen.. Bcs Zombies will come first then AI war
@Bhuvnesh right because someone is going to ask the AI, 'how to create a zombie apocalypse? '
yeah, and ChatGPT is the start
I love AI. Hopefully AGI will throw every Human from every position of power. Organic fucks, have been ruining plenty enough lifes over something as irrelevant as money. land and ressources.
Wow
It seems to me like it would make a better virtual assistant than it would a search engine. Sometimes I like having multiple web pages so I can compare ideas on topics. Or perhaps I’m looking for a specific web sight and not the answer to something.
Hey, whats wrong with one book with all the answers! ;)
@@ssamout memory. Forgetting the word by the time I have flipped to another page. Need 2 books side by side.
That's what Bing + ChatGPT is going to do. Bing will spit out web pages based on your conversation with ChatGPT.
@@ssamout the problem is to have a biased and sometimes wrong book as the source of truth
That's the beauty of learning and researching, and making informed logical decisions.
This is so amazing. I truly cannot wait to see how this AI war unfolds!
The beginning had me teary though. Just the innocence and the track doesnt make things better.
WHERE DID WE GO WRONG?!?
I am looking forward to this, we really need a search engine that finds what we are looking for instead of trying to sell us whoever paid the most. Something as simple as researching a specific car model, will likely show you a competitor car first, rather than what you are looking for.
Who's to say the ai wont be manipulated to give you a manufactured answer to the highest buyer?
I'm not convenienced.... the best answer may still take more then a simple search
I don’t think this will escape the claws of capitalism…
I guess youll be happy for the first couple years or months, then when the balances settle will the advertising on these engines ascend to a whole new level
@@Matanumi It's already censored/manipulated, try to ask it for a joke on a controversial topic, or something illegal (not necessarily illegal in your country).
I’m pretty confident that this ends up being one of those times where, no matter who wins, we lose.
There is always a risk. But I think people have the same fear about the Internet and how it affects books.
@@insertnamehere8625 It's not a worry of replacing old technology. It's a worry about the control of information at the behest of a massive corporate entity.
I mean it's *inevitable* isn't it??
AI will eventually become smart enough to conclude that 9 Billion humans is an unsustainable amount for Earth to support and all hell will break loose‼️
@@ban_tik_tok God, that would be so sick.
I, for one, welcome our new android overlords.
I think it's the opposite, this competition will be good for the consumer and for technology. Google will actually need to fight for its user base once again instead of only focusing on monetizing it.
The lists of search results that may or may not be what you want is irritating. This can be refreshing.
We'll have to see and I'm on the wait list
wow, I'm super excited about all this news regarding AI, looking forward to end of March
It will be interesting to see how companies monetize these AI-based search models. Also, such models will severely impact traffic on most websites, since most people will not go to the source website anymore. This will open up a whole set of new problems. Interesting times ahead and looking forward to more of your content on this topic!
It is very unlikely any company would fund any AI which could generate a result to a query which is detrimental to that company. Look at how the MSM will completely ignore a story which might affect one of their major advertisers. A certain group of drug companies comes to mind.
What I love about chatGPT is that it's minimalistic, you ask a question, you have an answer. Nothing in between. I'm afraid that Bing will integrate some kind of targeted ads with an IA using personal information that Microsoft collected. Or maybe it will answer in a way to push you to buy a specific product.
I wonder if we may have to pay a monthly subscription to the most useful versions of AI chat/search, that way there is no weird monetization scheme to mess things up
@@jonathaningram8157 When I search something I want many different answers. Giving just one answer is a problem. ChatGPT also seems very politically biased. Try to tell it to write a poem about Biden and try the same with Trump. It will give you a very interesting result.
@@johanhallgren While that is true for many things, there are things that you might need one answer for. Maybe you are searching for an established fact like "first exoplanet discovered", or "formula for something", etc. I can see it being very useful in that sense. In fact I'm already using it extensively (as in testing it) and I am generally VERY impressed. Like I ask it to compare 2 processors or phones and it lists down their specs and a general review (there might be some bias in the reviews maybe) and it gives me a very neat answer.
CharGPT is basically every futuristic game we played with AI all-knowing companion
The hang up is that the time it takes to have Ai get and have information ready in real time or even current data is not in the ability of Ai.
Bring on the AI competition, we will all benefit from these rapid technological developments.
The idea behind the terminator movie keeps getting closer and closer to reality.
That's no joke, China and USA made a small figure that melts and reassemble itself using magnetic fields. It's a small step, but decisively in the direction of the T-1000.
On the one hand, it's nice that Bing will be its own thing now, instead of just an inferior product or a mere Microsoft frontend for Google search.
On the other hand, the safety and bias issues are going to be trouble. AI sometimes produces *insane* results, and when you search for info, you need better than that. We're all impressed when AI works, but that success bias hides that potential for catastrophic failures. The AI art that doesn't understand how humans eat, or places children inside BBQ grills, is just the start.
I predict the next wave of "get onto the front page of search" goons will focus on ways to sneakily glut the AI's training data with your own product, politics, beliefs, etc. A new war.
Why is nobody talking about this?
---> All current biases present within the training data set instantly become "baked in" to CHAT GPT's AI algorithm... only to be repropagated at an exponential rate moving forward.
This is the ultimate social polarization tool... and the sooner people realize this the better.
Google: its like asking librarian and he gave you a book suggestion.
Bing(AI): its like asking something that know everything and itll gave you the Answer as detailed as you want.
The only downside to ChatGPT (in its current state) and the reason Google should be fine for the time being, is that Google offers live and up to date information. ChatGPT has to be trained over large datasets at a time and only knows up to a certain date (I believe atm 2021). This is not going to be particularly helpful things like "What time does X Coffee Shop close?" or "Find the nearest Target next to me" etc. Things could obviously change though in the near future.
Also add in the relatively high computational overhead: won't be cheap to run a global service based on it with current compute resources
I mean the database gives it the ability to analyze text and form a response, the database wasnt stored into the neural network were it just 'recalls' information in it and gives it to you, instead, the database was used in order to teach it how to read and process information and the human language, it just gave it the ability to 'understand' and 'formulate'. For finding up to date events it just has to have access to the internet, where it can use its neural network to formulate questions to, lets say, put into google, read a couple of the answers using, and spew out a few summarized sentences to you based on all the results it read.
Well, integrating time or location sensitive information into ChatGPT should not be a particularly insurmountable problem to solve.
I have found chatgpt fails to check knowledge bases being “confidently wrong “ even when challenged. Once a link is provided it says that it will update its records.
It’s clearly not live checking on it’s answer and if there is a change since 2021 it will probably be wrong. That’s why we are invited to correct it, getting and giving feedback is essential to a successful AI
Chat GPTs Chrome extension accesses the live internet
Timeline
00:00 Google reviews
02:07 ChatGPT reviews
03:25 Microsoft's Golden Goose and Google's Nightmare
08:40 More Problems for Google
10:34 How Google Could fight back
13:20 How Google Transformer gave to CHATGPT and Why Google never released consumer AI products.
14:45 From search Engine to Knowledge Assistant
Thanks!
Thank you for the timeline.
Thank you!!
@@Siscon92 no
modern hero
I would mutch rather find information through forums than an "Answer" from AI that can be doctored to give the "correct" Answer
It is too early to replace Google.
I think one problem with AI is it gets trained on historical data that can be 6+ months old. This can be a problem for a lot of searches, for example, "Where is the nearest [cuisine] restaurant to [location]?" might tell you to go somewhere that's been closed for a while.
Google search/maps are not updated either. This is due to people not updating websites
@@rerikm Why update websites when you don't get traffic, because the AI is giving the answers and entertainment?
I think a bigger problem is it stifles progress. For example there is a simple chemical cocktail, 2 cheap chemicals, that mixed together and taken diluted with water, can cure many modern illnesses. I've been online for decades and only recently discovered this; there is ZERO chance any AI would tell me about this obscure cocktail, that has already been censored on YT (and I'm not talking about ivermectin; this is old stuff discovered at least 12 years ago). Such discoveries are part of what makes the net so great, but all such things will be shoved down the memory hole, as we're fed synthetic knowledge that conforms to Wikipedia's bias and sponsorships.
Thats a problem for ChatGPT, not AI as a whole
Happens all the time even now.
Bing, generate a ColdFusion video for me that won't get me depressed about future.