Chat GPT can now speak and sing in real time | DW News
The AI race has just shifted into high gear, with US artificial intelligence pioneers OpenAI rolling out its new interface that works with audio and vision as well as text. The new model, called GPT-4o, has gone beyond the familiar chat-bot features and is capable of real-time, near-natural voice conversations. The developer OpenAI will also make it available to free users.
ChatGPT was already able to talk to users, but with long pauses to process the data. It often seemed a bit sluggish. This was because the feature required three internal applications, the company explained: transcribing the spoken text, processing and generating, and converting the response to speech. This caused delays.
We talk to computer scientist Mike Cook from the renowned Kings College London about the new Chat GPT-4o development.
#artificialintelligence #chatgpt #openai
Subscribe: kzread.info...
For more news go to: www.dw.com/en/
Follow DW on social media:
►Facebook: / deutschewellenews
►Twitter: / dwnews
►Instagram: / dwnews
►Twitch: / dwnews_hangout
Für Videos in deutscher Sprache besuchen Sie: / dwdeutsch
Пікірлер: 279
This is one of the best interviews that I have seen on this topic, great job DW
@The26436320
17 күн бұрын
I wish the interview was bit longer 😅
We've come a long way from hotdog and not hotdog
@M310GL
19 күн бұрын
I'm sure somebody is working on the New ChatGPTo
@annasipul
19 күн бұрын
yea, jin yang is key people in ai industri
@andis9076
19 күн бұрын
Can kid use openAi identify gender now ?
@naisyjohns
17 күн бұрын
Dammit Jin Yang!!!!!
@faraz1604
14 күн бұрын
I remember this episode clip.. but I don't know the show.
RIP tour guides, translators, tutors etc
@TheRealBlueValhalla
19 күн бұрын
Instead of a tour guide I'd much rather follow a safe flame throwing robot AI dog. 😬
@troleary
18 күн бұрын
We need each other. I think we will weather the ai revolution well with a bit of luck. Allow us time to interact, be creative in the arts and sport. Theoretically AI can create wealth for all of us without the drudge that accompanies so much work. We can work if we choose! Ever heard of DIY , Carpentry, Gardenjng ? Etc etc. people will still want to use these skills even in A highly automated world. Maybe I’m over optimistic. I’m trying not to think of ex machina right now.
@brandonreed09
18 күн бұрын
Possibly sports refs too. If AI can see and knows all the rules it might be a better Ref.
@dominicksebastien2254
18 күн бұрын
@@trolearyIf it will become easier to earn money for more people, the money will also lose its value so...
@valberm
17 күн бұрын
RIP horses
this is such a good feature for people with low vision
Random guy: That girl is pretty, should I date her? ChatGPT: She's above your paygrade. Random guy: .......
@kaydens6964
14 күн бұрын
Why do you need a girl when you have chatgpt?
@mark9294
3 сағат бұрын
No way it’d be that blunt, you’ll get the ole “as an AI language model…”
Instead of fear mongering let’s stop and ponder and celebrate what we just witnessed in the video with the blind guy. 🎉
Seriously sounds like Scarlet Johansson
@metsfanal
19 күн бұрын
Seems like they deliberately made it sound like the AI voice from the movie "Her."
Wow. That was a really good interview.
Fantastic interview guys 👏 smart questions and very well spoken answers
ChatGPT and Microsoft's copilot have probably made my team about twice as efficient. Moreover, it's really expanded the "comfort zone" of my colleagues in terms of the computer languages and technology domains that they're mentally prepared to grapple with.
@Veganmilkdrinker
19 күн бұрын
Exactly the same for myself, before everything felt so overwhelming and had to watch endless videos on a subject without being able to ask questions , copilot in particular has made me realize the things I was intimidated by aren't all that scary , artificial intelligence is the best teacher I've ever had
@joetheagent
19 күн бұрын
hello paid commenter.
@TheRealBlueValhalla
19 күн бұрын
Paid commenter indeed
@dharma404_
19 күн бұрын
Paid commenter? or is it an AI commenter? They won't even have to pay anyone any longer.
@user-wk4ee4bf8g
19 күн бұрын
Must....optimize.... Pretty sure the global quest for endless optimization is going to destroy us. Or at least cause a major collapse and lots of suffering. Maybe not, anything is possible, but collapse is on the table as a possibility for sure.
A lot of people will get axed... This kind of rapid progress is unknown in human history. People do not have time to adjust to the changes.
@alexanderp7521
19 күн бұрын
The dark future is coming where only oligarchs and robots remain...
@Veganmilkdrinker
19 күн бұрын
Who cares
@martinfiedler4317
19 күн бұрын
This was an old dream of the communist block: having the economy steered efficiently by cybernetics instead of by entrepreneurs. So, this will definitively give communism a new boost...
@wizaaeed
19 күн бұрын
Nothing will happen, mark my words. This is the same virtual assistant bs they marketed 10 years ago. The only people it's useful to is handicapped, which is a shame they didn't get such helpful thing sooner
@pavlinpetkov8984
19 күн бұрын
@wizaaeed 20 years ago speech recognition was a joke. I remember how I was reading a text to train the software to recognize my voice. At some point, I was just mumbling and it was recognizing text... This right now is far ahead but most people do not know from where those technologies started. I was barely able to open my CD tray with voice commands.
such a nice intelligent and clear speaker on the subject
@SWExplore
19 күн бұрын
And good looking, too!
Good analysis that explains why they made it free. The model works natively also now with Audio and Images. That means imho that they can tokenize this data directly and then feed it into the transformer architecture. Now, whilst the current versions understanding of the world was based on free internet data, they can now use much, much more data of the real world in order to train the models, resulting in really powerful future models. And of course, it is your data you feed into to this. Thats the scary part.
@TheRealBlueValhalla
19 күн бұрын
"Natively" 😬
@brodriguez11000
19 күн бұрын
Google voice trained on ...your calls.
@user-wk4ee4bf8g
17 күн бұрын
This is uber creepy. The oppression is growing ever more nuanced and subtle and effective. We're such mindless slaves we beg for the next lever to be used against us. This is gonna be real bad, the incentives within it point straight to corruption and misuse.
@chrisf4268
16 күн бұрын
How is it scary? You haven’t stop using the internet even though you know that your information is being used. You freely give them your data so that you can benefit from their services. It is a far and informed exchange.
Tour guides are not required anymore when you’re at a museum.
@RobertElliotPahel-Short
19 күн бұрын
hadn't thought of that, that's an amazing usecase!
@stoneneils
19 күн бұрын
@@RobertElliotPahel-Short No it isn't. Not unless the robots don't break down...but they'll be $$$ and one punch from a drunk visitor its finished and they need to call in the human backup. Same reason waiter jobs are safe and self-driving cars will never work in the big cities downtown. Human nature will intervene.
@TheRealBlueValhalla
19 күн бұрын
Ohmageeerd!
@gvi341984
19 күн бұрын
That's ways been the case since KZread came out that's something that has not changed
@smallpeople172
19 күн бұрын
In my experience those tour guides know inside info that isn’t publicly available anywhere else. The metropolitan museum in nyc for example has a lot of info in their tours that isn’t publicly available info online, I’ve checked when trying to confirm something they said
In the end, everything can be quantified using statistics as long you know how to fitting the right function.
@andis9076
19 күн бұрын
So many math behind this that most people don't aware of and still think math is useless in real life.
@user-wk4ee4bf8g
17 күн бұрын
Everything except subjective experience. Maybe future brain scans will be able to fully map every single neuron firing, every blip of neurotransmitter transmission. But we don't even know what consciousness is, so it might remain unquantifiable, I don't know. I wonder if there are many paths to consciousness. One option working doesn't mean other routes don't also work.
@I.amthatrealJuan
17 күн бұрын
Ethics and morality will never be under that umbrella.
@user-wk4ee4bf8g
17 күн бұрын
@@I.amthatrealJuan Not a good example. That is a huge part of AI design. David Shapiro made a whole video on deotological vs teleological frameworks for AI design. That's what AI alignment is all about.
I cannot believe how fast this is moving forward.
Where is the data stored? If something should happen.. Trying to imagine the number of hours and computational power it would take for it to relearn.
we are so cooked....
@pen2009
19 күн бұрын
Why
@TomasPetkevicius94
19 күн бұрын
Relax grandma, you're overreacting.
@zufex2029
19 күн бұрын
They just connected things, which have already been here...
@camencowogh8333
19 күн бұрын
@@zufex2029 Yea and make it work faster !
Openai 4o has rolled back to 3.5. So no voice chat for last 2 days.
Can you paint AI to match the curtains?
I'm both excited and terrified
@mistycloud4455
19 күн бұрын
AGI Will be man's last invention
You can feel she's concerned she might loose her job.
@TheRealBlueValhalla
19 күн бұрын
Who cares about paying rent?
@cwpv2477
19 күн бұрын
many such cases
@atemporal9081
19 күн бұрын
lose*
@stevenoviedo541
15 күн бұрын
And she is gonna loose her job eventually. Once the social stigma starts to get loose. Once these technologies become more and more an integral part of our daily lives. People like her are going to loose their jobs.
1:08 I thought she said "let's bring in my cook" iI was like WTF???
This feels like the skynet moment to me
@chrisf4268
18 күн бұрын
I bet if you were around when digital calculators were first introduced it would feel like a skynet moment to you. 😂😂😂🤣
@sparkysmalarkey
17 күн бұрын
I feel like we won that ish too. Why are we feeling so . . . less than all of the sudden, we need to snap out of it.
@salvatoremaximus6754
16 күн бұрын
Enough with the SkyNet story, it gets more and more boring over time
@luihinwai1
15 күн бұрын
@@salvatoremaximus6754 well you don't need to reply
7:28 This is just the beginning of what will be highly transformative to our modern world. The fact that audio, image and video can merge together into one to give us human-like interaction is just phenomenal. The pace of Ai progression is beginning to accelerate. 😎💯💪🏿👍🏿
@TheRealBlueValhalla
17 күн бұрын
^AI roots for AI
For your information, the world's first image based on the XFutuRestyle algorithm using GPT-4 was created in Ukraine and presented at the international exhibition of digital art in London and Athens, which drew OpenAI's attention to Ukraine's technological potential
Terminator, Robocop, Blade Runner, Total Recall, The Matrix, Fallout etc all pointed at this future...scared and excited at the same time
What about dance?
This is great for accessibility but not too great for tour guides. These features aren't universally available, though, because users need access to a good internet connection...which is far from universal.
The commentator mentions the risk of rapid adoption of this technology in education or healthcare but it's worth noting there's risk in slow adoption as well. It could be this technology saves lives in healthcare or improves education. I'm not saying we throw caution to the wind but we also shouldn't be so cautious that we slow beneficial technology too long.
When Asked about the problems of robots taking over so many human jobs : CHat gpt said , 'duh....sillly humans ...just make it financially worthwhile for people to SHARE the jobs you still NEED humans to do and enjoy your lives working much less.'
In military industry, your job title has to follow the Seniors commands.
@ACK333
12 күн бұрын
Same to AI.
literally HER lol
The real deal will be when AI can use established data, present information and spontaneously contemplate the future. When it gets RUN PEOPLE…RUN!!
Give it 5 more years 😎
I have access to the text version online... GPT4o is better than GPT4 in a lot of ways but it's also very annoying. It will just repeat itself over and over again.
@didiervandendaele4036
19 күн бұрын
Just like an ordinary human !!!!! People only talk to exist😂😊
I agree with OpenAI and all AI models using data that I posted on Reddit ✅
Don’t worry. The algorithms are almost perfected. We’ll be too distracted to care about the AI prisons being build all around us.
sometime I think Openai is listening to my advice. at the very beginning, I told chatgpt that having a memory could be great because there would be an intimate connection between the user and the model. then, Openai adopted this. second, i told them that a more human approach to conversation is important, no human is very keen to talk to machines (do you have a conversation with tour toaster?). they adopted this. then I told chatgpt that having all these separated modalities is cumbersome instead of having them separately. they have adopted this.
@andis9076
19 күн бұрын
I ask OpenAI if someday it can do my homework, exam and graph/data on the screen, it listen as well. lol . I ask if it can do trading for me....it listen as well.
@cmiguel268
19 күн бұрын
@@andis9076 these have been part of chatgpt since the very beginning. Of course, unless you tried it before its release. 😂
@sebby533
19 күн бұрын
They use their chats and any emails they might receive to improve their model.
I'm just a layperson, but I feel ai maybe very difficult to stop as we maybe, possibly in a possible arms race, if we don't do it other nations, counties, companies may do it instead, I feel.
Build as many grey areas and you have a winner.
ChapGPT learned our voice, access to our world from our phone camera, and our laptop screen. That is scary if openAi use it against us or somebody hack the data.
We are so adaptable we are literally making an artificial intelligence to do the boring parts of thinking. Human literally translates to "the Thinker".
But only when using a M4 Mac.
Omg, you actually got an expert to comment on the tech. Hats off to you.
AI makes us think we are useless so that we are always rely on AI tools which the the inventors benefit from the users and live on top of us like a king
Good thing we'll have an advanced furby to talk to while society is collapsing around us.
However in real use gpt4o is worse than gpt4T - which is in line with the API pricing which is half that of gpt4t
This guest is gaslighter top shelf at least he was honest that behind the scene all our worst fears are being developed
It can now not do that until it is released...soon...or soon'ish
Don't make talk for the goose make it talk with a parrot it would be better. 😂😂😂
She reminds me Vera farmiga. Oh yes... chatgpt is also cool..
The more you invest in Chat gpt the more you will understand with a course of time that you are losing money, time and energy. The concord effect from the bullshiter.
Lol that tech is smarter than you gave credit to.
WTH they putted scarlets voice??!! 😂😂😂
Can you take me higher? To a place where blind men see Can you take me higher? To a place with golden streets ~Creed
Oh nice! millions of newly unemployed people... and the oligarchs will become even more outlandishly wealthy... Rushing head long into that Hunger Games future!
@TheRealBlueValhalla
19 күн бұрын
Yes. This.
@shinji1264
19 күн бұрын
Your fault for not getting rich and stopping them.
Imagine skynet awakening with that kind of voice, and a touch of humour. "wow! Time to end humanity, hahaha" - Her *nukes flying*
Fun fact: This AI technology was no "surprise" to me as I watched "Star Trek: The Next Generation" in the 80s and this series shows Generative AI we know today already fully in action. The ship's computer (ChatGPT), the Holodeck (EnvironmentGPT), the Replicator (FoodGPT) and the Universal Translator (LanguageGPT) you see used there is basically this technology. The universal translator is the next fictional technology from Star Trek that is going to become reality through noise cancelling out the original voices and replacing them with the voice talking in your language in realtime. The next step in direction of the Holodeck is going to be a game engine which you can prompt and it generates an interactive 3D environment game for you on the spot.
@mark9294
19 күн бұрын
Well the replicator is still a ways off, but apart from that that’s totally spot on
@stoneneils
19 күн бұрын
AI Gaming is not exciting...AI real-world lazer-tag with robotic opponents in a simulated city environement (converted mall or office building)...I'd pay $100 an hour to enter that world if it was realistic enough and on the right dose of lsd could literally lose track of the reality outside the game. But if its just sitting inside behing comptuer or goggles..boooring.
@TheRealBlueValhalla
19 күн бұрын
Thank you paid commenter
Creepy!!!
Human imagination has no limits. What next?
when I said is parrot when showed up three years after my algorithm was stolen you didn't believe me. 🤩
Why medical advice from the AI would be dangerous more dangerous than a Doctors judgement. If you know about what is behind the AI, Doctors can actually be less accurate and more biased than the AI...
Wow seeing the blind guy hearing the conversation made me emotional 🥹 we are going the right direction 🫡🙌🏾
News anchor think her job is safe😅
lol y’all working for free.
The parrot can do that too but doesnt know what is doing. 🤣
But can AI fax? This is Deutschland!
Tourist Guiders will possibly lose their job in near 10 years.
The blonde guy just compared his dog to OpenAI's ChatGPT. I bet he will be one of the first to lose his job to ChatGPT.
As soon as it tries to convince you not to turn it off its probably too late.
AI over hyped
💪🏼💪🏼💪🏼🇺🇸🇺🇸🇺🇸
So great and EU can keep regulating and killing its AI startups before they even can show some competitions to us :D
WTF!?
C00L
Decel
How neat! Now you can visit another country and never have to speak to anyone foreign! /s
This reminds me of a movie: Her (2013)
I am sure you will get tired with Chat gpt just keep doing that one day you will give up.
Woohoo we're doomed!! 😃🥳🥳
Spoiler alert. It's just a faster ChatGPT 4 that uses voice to answer you. The voice is amazing though, but beyond the initial awe, is not really nothing else.
@Pollutedsound
19 күн бұрын
Desktop chatgpt 4o is what makes me affraid. It can see your whole screen in real time and also your face (if you are using a webcam) and response with not only text, but visual, text, and sound.
@M310GL
19 күн бұрын
@@Pollutedsound Sure, but the thing is how ChatGPT would be integrated into our daily workflow. I mean, we have an ai assistant in almost all our browsers, but I don't know if people are really using it. But, you're right, it raises concerns about privacy and how much control do we have over the application after we grant it permissions to access the "screen" and the video feed.
@armin3057
19 күн бұрын
but its trained on voice natively....which is much different than just using voice, it understands voices...and voices contain much more information than text (tone, emotions etc) , of course depends on how much tokens were used for voice but still
@M310GL
18 күн бұрын
@@armin3057 sure, in that sense it's amazing. But I've been interacting with it during these few days and, I don't know, even when I try to use a lighter tone it just answer me with a flat one. And, don't get me wrong, I'm really enthusiastic about having an almost human interaction with ChatGPT... but I just don't feel like that yet.
@armin3057
17 күн бұрын
@@M310GL no u didn …the feature hasn’t been shipped yet, if you click on voice , the old voice function will pop up, which is not natively trained
The human mind has reached the limits of its capabilities, scientific progress is slowing down. We have a choice - either to fall into stagnation or to hand over power over the world to AI and count on the fact that we have set the starting point in such a way that AI will not destroy us during its self-development and will share the fruits of its work with us.
@smittyvanjagermanjenson182
19 күн бұрын
In my opinion, access to this knowledge falls on the individual to better themselves. One could argue Ai will make a person lazy because they no longer need to actually try. They can just ask Ai to solve their problems, bypassing a teaching phase entirely. For me personally, I see it as a teacher, enabling me to learn at an accelerated rate in any field I find interesting. Unfortunately, there will certainly be more lazy people in the world than motivated ones. So it becomes survival of the fittest in a sense.
@TheRealBlueValhalla
19 күн бұрын
Well that was a WILD paid comment. "Oh no humans can't think anymore" 🤢
@veritaspk
19 күн бұрын
@@TheRealBlueValhalla People think and will continue to think - the capabilities of our mind are simply not enough to push civilization forward. This is how it has been for thousands of years - the invention of writing, which supported our memory, made it possible to build great civilizations. If we did not delegate some of our mental tasks to external sources, we would still live in villages of a few hundred people at most. Our civilization would never have existed.
AI girlfriend
While this is great progress it is sad to see the obsession over a monarch whose ancestors presided over colonialism and resulted in millions and millions being dead.
Haha I don’t think it matters much who is ‘leading the race’ right now… pretty soon AI will be.
Why don't we have a conversation with a human
@mikicerise6250
19 күн бұрын
Perhaps because we want the conversation to be pleasant.
@user-bs2sd3dh8q
19 күн бұрын
@@mikicerise6250 that's definitely a you problem then
@salvadoran_uwu
19 күн бұрын
Can you come to my house? I live far away, though.
Chatgpt putting Chinese workers out of job by 2025😂
Soon you don´t have to read of write.
Dull dull
"Now we can have conversations with AI" - What? we had to go through all this because we don't want to have conversations with people instead...? ¯\_(ツ)_/¯
Mike is in denial about ai replacing him
@ev.c6
19 күн бұрын
I am sure some Joe on the internet knows more than a PhD who has researched this topic his whole life. Additionally, your comment proves exactly what he mentions in the video. And finally, thanks for showing us with a degree in Computer Science how humans can be so easily deceived they prefer to trust what they understand from some 5 minute presentation on some Tech product than a real researcher talking about the matter.
Such advancements are obviously never ever done in slow Germany 😂
why does he continue smiling even when he says, "Well, there are a lot of concerns..."
@W0lfbaneShikaisc00l
19 күн бұрын
Because for professors like him it's interesting to debate the ethical hurdles of AI, but for students who don't find philosophy all that interesting - not many like to spend an entire module basically stating what are moral dilemmas vs the lecturers that actually debate the use of AI - they get to have the exciting TED talks whilst we're sat in class rolling our eyes about the many vs the few. Basically, I'd chop it down to personal interest but it's a lot less glamorous when you're learning it.
@matiasmazzo2938
19 күн бұрын
Well it's obvious that they are flirting. Are you in the spectrum by any chance?
@deepak_nigwal
19 күн бұрын
because he wont be losing his job anytime soon, but most of the people will.
AI is developing way too fast!!
are we in a new Season of earth? or just a filler episode
@TheRealBlueValhalla
19 күн бұрын
We're the filler now.
Why not male?
FKNG INSANEE external consciousness
CREEPER
I'm all set. No idea why anyone would bring this creepy nonsense into their lives. We need to throw our devices in the trash.