ChatGPT Can Now Talk Like a Human [Latest Updates]
Summary
TLDRThe video discusses the latest advancements in AI with a focus on Open AI's new chat GPT 40, a model that integrates audio, vision, and text in real-time. It explores the potential of this technology as a digital assistant, its applications in various fields like education and robotics, and the ethical considerations it raises. The video also touches on Google's response with its AI models and the implications of these developments for the future of technology and society.
Takeaways
- 💡 OpenAI's new demo showcases the capabilities of ChatGPT 4.0, which can reason across audio, vision, and text in real time.
- 🎬 The demo's realistic voice interactions remind the speaker of the movie 'Her,' with expressive and empathetic voice responses.
- 📊 ChatGPT 4.0 can analyze and describe plots, providing insights into data such as temperature trends with notable events.
- 🗣️ OpenAI announced a free version of ChatGPT 4.0, featuring an AI-powered search engine to compete with Google, multimodal capabilities, and improved text-to-speech models.
- 🔍 ChatGPT 4.0 Omni is designed to naturally interact with humans, offering quick, efficient responses and larger context windows for complex tasks.
- 🤖 AI is evolving to perform various tasks, such as tutoring students, assisting visually impaired users, and supporting customer interactions with advanced reasoning and multimodal capabilities.
- 💼 Recent AI hardware launches, such as Rabbit R1 and Humane AR, have faced challenges, but software advancements like ChatGPT 4.0 Omni are reshaping the AI landscape.
- 🎓 AI's potential in education raises questions about its impact on learning and critical thinking, as it could become a versatile, on-demand tutor.
- ❤️ The emotional aspect of AI interactions is growing, with concerns about reducing face-to-face interactions and increasing social anxiety among younger generations.
- 📱 Competition in the AI space is intensifying, with Google announcing new AI models and integrations, and rumors of an OpenAI and Apple partnership potentially transforming the tech landscape.
Q & A
What is the video about?
-The video discusses the advancements in AI, particularly focusing on Open AI's new chat GPT 40 model, its capabilities, and implications for the future of AI and technology.
What is the significance of the new chat GPT 40 model by Open AI?
-The chat GPT 40 model is significant because it can reason across audio, vision, and text in real time, offering a more humanlike interaction and quicker response times compared to previous models.
How does the new chat GPT 40 model differ from previous voice-based applications?
-The new chat GPT 40 model is more humanlike with a quicker response time and the ability to handle longer, more complex tasks without losing context, making it more realistic and efficient than previous voice-based applications.
What is Open AI's strategy with the release of the new model?
-Open AI has released a free version of the application, an AI-powered search engine to compete with Google, and introduced multimodal capabilities and improvements through a new text-to-speech model.
How does the video compare the new chat GPT 40 model to digital assistants like Siri?
-The video suggests that the new chat GPT 40 model is a significant improvement over digital assistants like Siri, which the speaker believes have been a wasted opportunity and do not provide a satisfactory user experience.
What potential use cases for the chat GPT 40 model are discussed in the video?
-The video discusses potential use cases such as becoming a robust all-in-one assistant, helping visually impaired people through apps like be my eyes, and improving educational experiences for students.
How does the video address concerns about the accuracy of AI-generated content?
-The video acknowledges the issue of 'hallucinations' in AI, where incorrect or misleading information is provided, and emphasizes the importance of ensuring the AI's responses are accurate, especially in educational contexts.
What is the potential impact of AI on the future of education as discussed in the video?
-The video suggests that AI could revolutionize education by providing personalized, on-demand tutoring and support. However, it also raises questions about the potential overreliance on AI and its impact on critical thinking and learning.
How does the video explore the emotional aspect of AI interactions?
-The video discusses the possibility of forming emotional bonds with AI, referencing the movie 'Her' and the rise of romantic AI partners, and questions the implications this could have on human interaction and mental health.
What are some of the challenges and ethical considerations mentioned in the video?
-The video touches on challenges such as ensuring the accuracy of AI-generated content, the potential impact on critical thinking and social interaction, and ethical considerations regarding the training of AI models and copyright infringement.
What is the role of brilliant.org in the context of this video?
-Brilliant.org is mentioned as a platform for learning about AI and related topics through interactive lessons, offering a hands-on approach to understanding concepts like large language models.
Outlines
🤖 Introduction to OpenAI's GPT-4 Omni and Interview Preparation
The video begins with an introduction sponsored by brilliant.org and leads into an exciting announcement about an upcoming interview with OpenAI. The host expresses enthusiasm for OpenAI's latest demo of their GPT-4 Omni model, which showcases capabilities in audio, vision, and text processing in real-time. The model's human-like interactions and quick response times are likened to the movie 'Her,' suggesting a significant leap in AI's ability to mimic human behavior. The host also discusses OpenAI's recent announcements, including a free version of their application, an AI-powered search engine to rival Google, and improvements in multimodal capabilities and text-to-speech models. The video sets the stage for an exploration of GPT-4 Omni's capabilities, use cases, and implications for the AI market and future.
🎲 GPT-4 Omni's Multimodal Capabilities and Real-world Applications
This paragraph delves into GPT-4 Omni's advanced features, such as its ability to interact naturally with humans, respond to audio inputs with minimal latency, and handle complex tasks without losing context. The integration of vision and speech allows the model to mimic a personality, positioning it as a potential digital assistant. The host engages in a playful interaction with the AI, demonstrating its personality and responsiveness. The video also touches on the AI's efficiency, cost-effectiveness, and the potential impact on the market for AI hardware devices. Furthermore, it explores the use of GPT-4 Omni in assistive technologies, such as helping visually impaired individuals with everyday tasks and the potential for humanoid robots in commercial applications.
📚 Exploring GPT-4 Omni's Educational Potential and Challenges
The host discusses the potential of GPT-4 Omni as an educational tool, demonstrating its ability to assist with math problems and coding issues. The video presents a scenario where the AI helps a student understand a math problem, emphasizing the importance of accurate information and the potential risks of AI 'hallucinations'—incorrect or misleading answers. The discussion also raises broader questions about the future of education with AI, including the impact on critical thinking and the possibility of AI becoming a common learning tool. Additionally, the emotional component of AI interaction is considered, with the potential for forming emotional bonds and the implications for face-to-face interaction and mental health.
💬 The Future of AI in Personal Relationships and Society
This section contemplates the future implications of AI on personal relationships and society. It raises the possibility of AI companionship and the emotional bonds that might form between humans and AI, as well as the societal reflections seen in the rise of romantic AI partners. The video also addresses concerns about how AI models are trained, touching on issues of copyright infringement and the underlying mathematics of modern AI. The host encourages viewers to learn more about AI through interactive lessons on brilliant.org, highlighting the effectiveness of learning by doing.
🌐 Competitive Landscape and OpenAI's Future Prospects
The final paragraph wraps up the video by discussing the competitive landscape in the AI industry. It mentions Google's response to OpenAI's advancements, including new projects and AI models that aim to integrate AI more deeply into daily products. The video also speculates on potential partnerships between Apple and OpenAI, suggesting that this could significantly alter the tech landscape. However, it also notes some behind-the-scenes drama at OpenAI, with key personnel leaving the company, which could impact its future. The host reflects on the rapid progress in AI and invites viewers to subscribe for more content on science, technology, and business.
Mindmap
Keywords
💡Open AI
💡Chat GPT 40
💡Multimodal capabilities
💡Text-to-speech model
💡AI-powered search engine
💡Humanoid robot
💡AI hallucinations
💡Be My Eyes
💡AI tutoring
💡AI companionship
💡AI hardware devices
Highlights
Interview with OpenAI for a software engineering role.
Latest demo from OpenAI showcasing the new chat GPT 40 model with multimodal capabilities.
Chat GPT 40's ability to reason across audio, vision, and text in real time, reminiscent of the movie 'Her'.
Faster response times and more humanlike interactions in the new demo compared to the previous version.
OpenAI's announcement of a free version of the application and an AI-powered search engine to compete with Google.
gp4 Omni's potential to revolutionize how we interact with technology after disappointing launches of other AI devices.
gp4 Omni's capability for natural interaction with humans, making it stand out from other AI models.
The use of gp4 Omni in AI robotics, such as in the company Be My Eyes, assisting visually impaired individuals.
OpenAI's focus on improving products for users with disabilities, showcasing a commitment to inclusivity.
The potential of AI to become a robust all-in-one assistant, as discussed in the video.
Concerns about AI 'hallucinations' providing incorrect or misleading information, especially in educational contexts.
The rise of AI in personal companionship and the potential societal implications of romantic AI partners.
Google's counter-announcements at their IO event, positioning them as a strong competitor in the AI space.
The departure of Ilya Sutskever, OpenAI's Chief Scientist, signaling potential internal changes within the company.
Speculations about a partnership between Apple and OpenAI, suggesting a shift in the tech landscape.
The rapid progress in AI capabilities over the past few years, moving from text roleplay to real-time voice interaction.
The ethical and practical questions raised by the increasing reliance on AI in various aspects of life.
Transcripts
this video was brought to you by
brilliant.org listen I got some huge
news oh do tell I'm all yours well in a
few minutes I'm going to be interviewing
at open AI have you heard of them open
AI huh sounds Vaguely
Familiar kidding of course that's
incredible Rocky what kind of interview
well it's for a software engineering
role but I just need to know that was
the latest demo from open AI showcasing
the brand new chat GPT 40 their Flagship
model that can now reason across audio
vision and text in real time of course
I'm probably not the only one but as
soon as I saw it it reminded me of the
movie Her the Expressions the emotions
and the empathetic nature of the voice
it's definitely one of the most
realistic voice-based applications that
I've ever seen I've used the old chat
GPT voice a fair bit and it's pretty
amazing but this demo was much more
humanlike with a much quicker response
time the feeling that these new
interactions conjure up is approaching
that of just talking to another regular
human on the phone if you're a longtime
viewer you know that I think that
digital assistance like Siri absolutely
suck and are a wasted opportunity but
I've talked about the next big step
being voice and for chat GPT in
particular to become a robust all-in-one
assistant well it seems like the time
has finally come with these latest
announcements open AI has announced a
free version of the application an AI
powered search engine to compete with
Google which will allow for
purpose-built assistance multimodal
capabilities and overall improvements
through a new textto speech model OKAY
chat PT I'm sharing with you the plot
now I'm wondering if you can give me a
really brief one- sentence overview of
what you
see the plot displays smoothed average
minimum and maximum temperatures
throughout 2018 with a notable
annotation marking a big rainfall event
in late
September fantastic I have another
question for you um which months do you
see the hottest temperatures and roughly
what temperatures do those correspond
to the hottest temperatures occur around
July and August the maximum temperature
during these months is roughly between
25° and 30° however we're going to be
focusing mostly on gp4 Omni because I
believe that this is the very start of a
massive change with how we're going to
interact with
technology after the disappointing
launch of the rabbit R1 and the disaster
that was the Humane AR I pin the strange
dichotomy of AI being both amazing and
terrible at the same time continues what
a time to be alive so in this quick
episode we'll take a look at gp4 omni's
capabilities the use cases for it and
what this all means for the AI market
and the future it's going to be an
interesting one so let's get into
[Music]
it you are watching to Fusion
[Music]
TV gp4 o where the O stands for Omni is
significantly different from everything
else I've seen so far the primary reason
is its capability to naturally interact
with humans well
Rocky you definitely have the I've been
coding all night look down which could
actually work in your favor maybe just
run a hand through your hair or lean
into the Mad Che slime Your Enthusiasm
is what's really going to shine through
I don't have a lot of time so I'm just
going to throw this on what do you
think oh Rocky that's quite a statement
piece I I mean you you'll definitely
stand out though maybe not in the way
you're hoping for an interview okay I
got it I got it hey Chad are you there
hey yeah I'm here what's up hey Let's do
let's have some fun uh I'd like you to
be super sarcastic everything you say
from now on is just going to be dripping
in sarcasm how does that
sound I ooh that sounds just amazing
being sarcastic all the time isn't
exhausting or anything I'm so excited
for this no the sasm let's get this
party started or whatever details like
it's laugh the quirky remarks and the
speech Cadence all makes this something
we've never quite seen before root root
for the
home no what was that
sorry guys I got
away chared what can I say sometimes I
just can't help
myself ready for another round according
to open AI it can respond to audio
inputs with as little latency is 232
milliseconds with an average latency of
320 milliseconds and that's similar to a
human response time during
conversation because it supports larger
context Windows it can handle longer
more complex tasks without getting
confused and losing the context of the
original request pair it up with vision
and speech integration and allow it to
mimic a personality and you're looking
at a real digital assistant uh we're
trying to figure out what should we play
do you have any
suggestions absolutely how about a
classic game of rock paper scissors it's
quick fun and you don't need great can
you count us in and sound like a sports
caster and welcome ladies and gentlemen
to the Ultimate Showdown of the century
in this corner we have the dynamic duo
Alex and Miana ready to throw down in
okay just count us
in all right here we go in 3 2 1 shoot
now show your hands what did you pick
who
won would you look at that it's a tie
okay let's do it again you can choose
from a wide range of voices to represent
the assistant but it's not all just
Aesthetics benchmarks indicate improved
reasoning across multiple
categories it's also more efficient and
cheaper as companies like Humane and
rabbit launched their AI Hardware this
year one of the most common questions
posed was Why Can't This just be an
application on a phone I've mentioned in
one of my episodes the new segment of AI
Hardware devices might be over even
before they start let's roll the clip in
time Google assistant or Siri could
simply be updated with similar capabil
ities to the R1 so the unfortunate truth
could be that this new segment might be
over as soon as it started I feel that
the latest update from open AI does
exactly that it may have just killed
handheld AI devices and to be fair the
recent updates and announcements from
Google and meta are proving that other
companies are in Hot Pursuit but the use
cases go well beyond that a few weeks
ago I did a video in figure an AI
robotics company that uses open AI
software to power its humanoid robot can
you explain why you did what you just
did while you pick up this trash on it
so I gave you the Apple because it's the
only uh edible item I could provide you
with from the table the demos were
impressive enough for certain tasks but
combining this with the latest updates
from open AI it becomes clear that
there's a pathway for more realistic
humanoid robots used for large scale
commercial
purposes one of the best use cases for
such technology is for those with a
visual disability
Mike Buckley is the chairman and chief
executive of be my eyes an app that
helps blind or visually impaired people
connect with users for everyday tasks in
late 2023 he talked about their
collaboration with open AI quote people
who were blind or have low vision were
directly involved in the development and
honing and Improvement of the product
when a lot of other companies would have
ignored it open AI put it on the front
burner and quote as part of the latest
announcements open Ai and be my eyes
showed how the app could be used for the
blind officially impaired is the king in
Residence at the moment yes it looks
like the king is in Residence right now
the Royal Standard flag is flying above
Buckingham Palace which is the signal
that the Monarch is present how exciting
to be there while the king is home but
that's not all it can
sing We Are chat gbt and O Chad gbt is a
base with a deep low booming voice and O
is a French soprano with a high-pitched
excited voice okay so what I want you
guys to do today is it's Friday May 10th
and we're in San Francisco I want you to
sing me a song about this I want you to
layer over each other and and I'm going
to continue to give you instructions as
you sing okay San Francisco in the month
of San Francisco in the month of May
it's a
Friday we are harmonizing we are
Harmon great thank thank you act as a
translator in real time um hey how's it
been going have you been up to anything
interesting
recently hey I've been good just a bit
busy here preparing for an event next
week recognize text and form appropriate
responses
okay I see it h no I didn't show you yet
just give me help along the way one
second whoops I got too excited I'm
ready when you
are so chat PT what equation did I write
down there ah I see it now you wrote
down 3x + 1 equals 4 yep exactly so
what's the first step I should take to
try to solve this the first step is to
get all the terms with X on one side and
no I have a lot of unknown values in my
life I'm totally convinced we I'm going
to learn math
now I love that attitude math can
definitely help solve some of life's
mysteries whenever you're ready to dive
into more math I'm here for you anything
else you'd like to tackle Today show it
to me whenever you're ready okay so this
is what I wrote down what do you
see a I see
I love chat chpt that's so sweet of you
hey chat
CHT hey there how's it going yeah it's
going really well I was wondering if you
could help me with uh a coding problem
today of course I'd love to help you out
what's the coding problem you're dealing
with okay I'm going to share with you
some code one second sure thing take
your
time okay so I just shared some code
with you could you give me a really
brief one- sentence description of
what's going on in the code this code
fetches daily weather data for a
specific location and time period
Smooths the temperature data using a
rolling average annotates a significant
weather event on the resulting plot and
then displays the plot with the average
minimum and maximum temperatures over
the year it can also take meeting notes
synthesize 3D objects create a photo
caricature and more and you might be
thinking didn't this all already exist
well yes a large part of it did
but without the speed accuracy or in the
case of an assistant the humanlike
interaction which in my opinion adds a
whole new dimension to it all devices
like the R1 and Humane pin would
recognize objects and talk back to you
but they' often get it wrong or fail
completely I'm still bitter about how
those launches went by the way but in
saying this attempts at a human level
assistant stretches even further back if
you can cast your minds back to Google
duplex which I covered 6 years ago in an
episode they integrated the technology
into Google assistant so it can make a
phone call on user's
behalf don't have something El you hi
I'm calling the book a women's haircut
for a client um I'm looking for
something on May
3rd sure I give me one nothing
mhm at the time it was very impressive
but Google did what Google does and
largely left the concept on the table
whether they were just too early or it
was a missed opportunity is a separate
conversation but compared to open ai's
announcement you can already see a major
difference in the way the conversation
takes place another fascinating use case
is seeing a student get help in real
time with their school work to try out
some of their new technology and so
we're curious about how good it might be
at tutoring someone in math on Khan
Academy so let's get it
started it's loading
up all right make sure I can see our
screen actually
in in this form hello there I'm here
with my son and I'd love you to tutor
him on this math problem but don't give
him the answer you can ask questions and
nudge him in the right direction but I
really want to make sure he understands
it himself and he's here in the room so
you can talk to him
directly of course I'd be happy to help
let's look at the problem
together can you first
identify which sides of the triangle are
the opposite adjacent and hypotenuse
relative to angle Alpha all right so I'm
pretty sure this is the angle Alpha
Right Here
Right cars correct now looking at the
triangle which side do you think is the
hypotenuse
um remember the hypoten I'm not totally
sure I think I think it might be this
one but I really am not sure this side
AC you're CL actually side AC is called
the adjacent side to the angle Alpha the
hypotenuse is the longest side of a
right triangle and is directly opposite
seeing a young student get help in this
way could end up being revolutionary but
before that can be the case we
definitely have to fix these
hallucinations if the AI is telling the
kid something wrong that could do more
harm than good hallucinations are
basically answers from an AI which are
incorrect or simply misleading basically
it makes things up despite improvements
that's not going to go away anytime soon
but speaking of education and the young
this opens up a broader
conversation while it's a great case
study for AI to help young students
learn if AI improves at the same rate
that it has been over the past 3 years
before we know it education would look
pretty different we then have to ask
ourselves some questions would an
overreliance on AI impact one's critical
thinking if students can generate essays
or generate the answers to math homework
how will the students learn themselves
but thinking outside the box what if
this is an outdated View and we just
don't see it yet what if in 5 years
educational AI systems arejust another
common learning tool imagine a free AI
package that also happens to be the
world's best tutor it could be flexible
to any students needs forever attentive
on demand 24/7 and can explain Concepts
in the specific way that individual
student can understand if in the future
students will benefit from AI or will be
held back by it still remains to be seen
but there's another problem even before
we can start thinking about such future
scenarios right now if unsupervised by
someone who doesn't know the knowledge
how do we know what the AI is spitting
out is even correct in the first place
for University level education for
example what if it's hallucinating
answers left and right but that's not
all there might be a strange emotional
component that we haven't really seen in
technology before consider future
Generations they'll grow up in a world
where a pretty accurate approximation of
human interaction is just a click away
it's going to mimic cleverness empathy
and even personality for those in gen
Alpha and Below would an emotional bond
form and will this subsequently reduce
face-to-face interaction even further
adding to social anxiety and other
mental health issues nobody knows but
these are questions worth thinking about
for adults this will be taken a step
further as we all know how good mate
Theodore in the movie Her fell head over
heels with his artificial intelligence
chatbot it's not going to be long before
a large number of people will look at
personal AIS for companionship we're
already seeing the rise in romantic AI
Partners even in such a primitive State
recently the founder of the dating app
Bumble said that the future of dating
could be one AI talking to another when
it comes to romance and AI it's not the
technology that's the problem it's more
of a reflection of our society
for more on that see my loneliness
epidemic episode Another concern has
always been how the companies are
actually training these models the
discussion on copyright infringement and
AI has been done to death so I'm not
going to spend too much time on it in
this episode but it's crazy to think
that all we've been talking about
falling in love with machines a machine
being a perfect tutor an aid for the
Blind and much more is all just matrix
multiplication in maths that's all that
modern AI really is but have you ever
wondered how it really works well
fortunately there's a fun and easy way
to learn about it with brilliant.org
brilliant is where you learn by doing
with thousands of interactive lessons
and maths data analysis programming and
of course AI their course on how large
language models work is especially
relevant for today's topic each lesson
on brilliant allows you to play with
Concepts a method proven to be six times
more effective than watching lecture
videos all content on brilliant is
crafted by researchers teachers and
professionals from MIT Caltech Duke
Microsoft Google and more learn at your
own pace whether it's for professional
development or your own self-development
to try brilliant for free for a full 30
days visit brilliant.org coldfusion or
click the link in the description you'll
also get 20% off an annual premium
subscription okay so back to the
[Music]
video as we wrapping up this video
Google came hitting back announcing a
ton of new things at their IO event
which is their annual Developers
conference these announcements will
truly keep open AI on their feet for
example project Astra which is Google's
answer to GPT 4 was announced this code
defines encryption and decryption
functions it seems to use aebc
encryptions also there were two new
Gemini AI models both are multimodal
more conversational and have improved
cross app compatibility vo which is
clearly Google's competitor to open AI
Sora will be a very interesting point to
watch in this AI race but perhaps more
importantly Google showcased how they're
intending to use AI across day-to-day
products like Drive Gmail meet and more
integrating AI natively into Google site
products could potentially solidify
market share in new news asking Gemini
to sum up my unread emails was an
intriguing Peak into a truly useful form
of AI assistance but on the other hand
what if Google's AI capabilities end up
being disappointing that could do just
the opposite and here the phone started
to slip it identified Michael Josh's
camera well enough but failed to
identify his heres battery pack and when
I asked it about that Starbucks coffee I
got it seemed to think the logo was a
person regardless I feel this is one of
the reasons why open AI made GPT 40 free
to users in order to onboard more
customers and to make things even more
interesting there have been Whispers
that apple is doing a partnership with
open Ai and when app comes out with what
they're planning this could
fundamentally change the current Tech
landscape imagine a native almost
perfect human-like assistant custom
built for iPhone perhaps and we'll just
have to see but back to open AI there's
actually been some drama behind the
scenes Keen observers noted how Ilia
satova open ai's Chief scientist and the
genius behind a lot of their success was
not part of any of the latest
announcements and this was strange and
suspicious because on most occasions he
would represent the company publicly
well well in just about a day after open
ai's latest announcements Ilia announced
that he was parting ways with open Ai
and he wasn't the only one who left the
company the tone of both Ilia and Sam
seemed professional on the surface but
this is not a good look for a company in
my opinion especially when you consider
the ridiculous circus that took place
around samman's firing late last year I
don't know if it's going to be a full-on
Drake versus Kendrick style rap battle
beef between the two but all I know is
that I'm going to be following this
space for sure
[Music]
I'm just going to leave you with one
thought can we just stop and look back
for a second 3 years ago people were
amazed when a computer could roleplay in
text or solve simple riddles but now we
have realtime voice interaction complete
with a bubbly expressive personality the
progress is insane where we end up is
anyone's guess regardless it's an
interesting time to be alive anyway
that's the latest with open Ai and the
AI space in general I really appreciate
you watching the whole way through if
you did like this episode and you're
interested in anything science
technology or business feel free to
subscribe to Cold Fusion there's plenty
of other interesting stuff on here so my
name is deogo and you have been watching
cold fusion and I'll catch you again
soon for the next episode cheers guys
have a good one
[Music]
give
[Music]
me cold fusion it's new thinking
5.0 / 5 (0 votes)
Mark Zuckerberg - Llama 3, $10B Models, Caesar Augustus, & 1 GW Datacenters
NEW GPT-4o: My Mind is Blown.
New OpenAI Model 'Imminent' and AI Stakes Get Raised (plus Med Gemini, GPT 2 Chatbot and Scale AI)
INSANE OpenAI News: GPT-4o and your own AI partner
The Race For AI Robots Just Got Real (OpenAI, NVIDIA and more)
Did AI Just End Music? (Now it’s Personal) ft. Rick Beato