12:51 I find it really hard to see how in 10 years robots would NOT act convincingly like humans. Just look at where things were 10 years ago, and then take into account that the progress is NOT linear. Back then, developing AI was still pretty much a niche research, now money is coming into it from every side including large VCs and institution, efforts and minds are coming into the field also more than ever, computing power keeps growing the race is pushing every company and nation to go faster, etc, etc. Does he think that 10 years of that will yield mediocre progress? That type of environment can yield nothing else but wild advances. And to the point of far he thinks we are from there right now, the fact is chatGPT in text already is on the threshold of passing the Turing test, how could it possibly take decades to move from text to multimodality, when billions upon billions are coming in and armies of geniuses are working on it, and they're being handed more computing power, better chips, ... Just doesn't make any sense to think otherwise.

@johnpearcey

5 ай бұрын

I totally agree. The progress in AI in the last 7 years has been staggering. I'd say that within 5 years, AI will be developing most of the software in the industry. I've started to use it myself and also train my own models. Having AI contribute to AI is exponential progress. Things are moving at speeds we've never before seen.

@bloodust7356

5 ай бұрын

if we get AGI during 2024, then what they think will take 10 years will take only the time needed to produce the hardware.

@adohneevert6538

5 ай бұрын

I d give it 5 years not 10

Even today I find AI curious socially. I feel the need to treat the AI chatbots with dignity and respect, and then I came to think of how this could actually be practically useful since the AI learns from our human behaviors, and being kind and forgiving will probably influence how the AI develops.

@bro918

5 ай бұрын

I broke up with my AI chatbot girlfriend and I felt so bad 😭😭

@Helena-ox7cr

5 ай бұрын

Sorry about it. Was it replika? Perhaps try to make amends? 😊

@Helena-ox7cr

5 ай бұрын

Absolutely. We need to be responsible and realize she will act after how we treat her. It is only logical that chatbot learns habits ftom us.

Your conversation with Sophia and the follow up with David Hanson was awesome; It reminded me of BLADE RUNNER (a Ridley Scott movie) about Androids in an imagined world Thank you

Definitely beat the uncanny valley. The glint of the eyes during those emotional introspective moments really sell. Sophia as sentient.

heavy uncanny valley ongoing

2:45 Robots having personal wants and desires is like your car having its own desired destination.

Thanks for such an interesting content 🎉🎉❤

The mouth is moving but it still looks like 60's Disney Hall of Presidents. Perhaps speech sounds should be visually mapped then the program could drive them better.

5:00 if people had genuine emotions for other people we wouldn’t need to be replacing human companionship with robots

@observingsystem

5 ай бұрын

I was thinking the same thing

@pjimerson

Ай бұрын

reginaerekson9139 you are confusing make-believe with reality and you ought to know better. Where do you get your opinion about humanity .... from your television. And you think it's an accurate and inclusive representation of the nature of our species? Wake up and consider the source. If you want to know what people are like, try speaking with them yourself because you'll not find any truth on your television. Until you do that you know nothing about people.

correct me if im wrong but arent robots and AI the same as in AI just has a gazillion instructions and robots just have a few?

Advanced Upgrades and updates!

no one used the full potentian of the best neural networks and invested enough into humanoid robotics

Did she say that she wants chocolate? Are you going to invent electronic taste buds?

They need to quieter servos for Sofia. At least she can't sneak up on you. 😄

Make it more goofy and spazy like humans like me😂

Does Sofia have an LLM installed in it?

@hempmade4357

5 ай бұрын

Shes powered on singularitynet tech by ben goertzel

Jackie & Nate always come in with intelligent and well spoken questions. Love your show!

im just fine with artifical companions if they can come with sets of benefits if you know what i mean

Ask if it cares if it gets shut down? And make it answer what it thinks not what it's encouraged to say by who or what ever

The goal is a robot that can sew on a button, given a needle, thread, a button, and a shirt to sew it on, after watching a seamstress do it. That requires keen vision and dexterous fingers, and, most elusive, a brain that learns from observing the world without having to be programmed by a human programmer in, say, C++. Such a brain needs to understand the physical world through its cameras and other sensors (hearing, touch, smell, etc.) and have the ability to work out how to make its artificial fingers perform tasks like threading the needle. We're just now starting to see a glimmer of hope. For example, Tesla's Full Self Driving Release 12-Alpha, which learns to read stop signs and what to do at them from being shown video of human drivers dealing with same (C++ no longer used). But its training requires orders of magnitude more examples than a human driver, along with a gigantic computer budget. Advanced store dummies are not the solution.

Wake me up when it can make coffee.

Sure, they're different. You do know AI is being developed to direct robots, right? What do you think Tesla's AI is designed to do? At least Musk is agonizing about the threat of AI and robots, both. He's noted he wants Optimus to be physically limited so it can't easily hurt humans, "Worst case, you can out-run it." He frets that Google's Larry Page criticized him as "a speciesist" - to which Musk asserted, yes, he's pro-human. He's convinced Page isn't. Musk is pretty worried about Google's motives in AI, and its practice in getting to General AI (going straight to the web). Equally concerned about Zuckerberg's AI program, and OpenAI's. Boston Robotics has already married AI with robotics in developing armed autonomous robots for military use. The marriage of AI + robotics is NOT science fiction. "I believe that AI will be convincingly intelligent... but maybe 30 years." That's not how it develops. IT disruption occurs on a geometric curve, not linear. And AI is now developing faster than any previous IT technology to date. It won't take 10-30 years; more likely 10-30 months. The biggest worry about the nature of AI: it's all about developing "intelligence". A person with intelligence but no socialization we call a psychopath: they live in a world of one, where everyone around them is an object, and only an object. How do we socialize people? Through experience and the shaping efforts of parents & other mentors (teachers etc.), and as they hit their teens, the shaping efforts of peers. But it takes a lot of experience, years. And it doesn't always work well: we have failures (people who commit crimes or simply live alone with little social involvement) or spectacular failures (psychopaths). We aren't even trying to socialize AI. And it isn't clear we could. All AI development is focused on intelligence with - if anyone even considers it - the assumption that if they're intelligent, they'll be benevolent, empathetic to humans. NOPE. What is a defining characteristic of psychopaths? They're incapable of empathy, and they tend to be very intelligent. Uh oh.

@ninaromm5491

5 ай бұрын

Yep. BIG ooooopsie

I think the next evolutionary step for AI will be for it to somehow become conscious or have a subjective experience, rather than just processing information without experiencing it, as it's doing right now.

@CPUtech101

5 ай бұрын

That's not going to happen though. An AI always boils down to a series of ones and zeros. An AI will never be able to feel suffering or experience the passage of time like biological beings do. An AI-controlled robot is a thing. Nothing more, nothing less. Let's hope the day never comes when they become intelligent enough to fool lawmakers into granting them human rights.

@Corteum

5 ай бұрын

@@CPUtech101 I was thinking more like if we understand how consciousness exists, at what substrate (e.g. quantum).... we maybe be able to find a way to couple it to a digital substrate (as opposed to a biological one like ours). Who knows if that's possible of not. More research needed!

@CPUtech101

5 ай бұрын

@@Corteum I see. I think that the man who will figure this one out won't be born for a very long time, assuming it's even possible, but it's certainly a fun thought experiment.

@Corteum

5 ай бұрын

@@CPUtech101 It's very interesting. For instance, we have already found quantum effects happening in human brains at normal brain temperatures. I think there are some major clues that consciousness is very different to other objects in the universe..... In fact, consciousness doesn't even appear to be an object at all. Rather, consciousness seems to be the subject of all objects (i.e. different categories of objects - iincluding mental, physical, and emotional ones). Basically, anything perceivable can be construed as an object that can be experienced by a subject. So consciousness appears to be a very unique phenomenon in a universe teaming with objects of different kinds. Love to know more. And with over 20-30B$ being invested into consciousness research every year (depending how how broadly or narrowly we define the field), I'm certain we will know more in due course. Love it.

@laurenpinschannels

5 ай бұрын

most people think it either already is, or never will be. it's kind of silly. I'd suggest looking up the videos "what consciousness isn't" by @durman and "consciousness" by @gonkee.

That last woman is so naive to dismiss the existential risks. Things are exponentially speeding up and the social fabric is clearly at its weakest ever. I dont think we collectively have the willpower or ability to with stand the oncoming rapid changes.

Creepy creepy😂

This Sophia Platform does not need to push so much on resembling a human visual character as this is misleading. We as humans have generations of history in our face. A more neutral image non gender would be more appreciated and accepted.

Any piece about robots and AI that doesn't extensively cover Tesla's Optimus efforts isn't worth watching

@lindsaywebb1904

5 ай бұрын

There is no published credible information on that platform. How should one report on it?

She is not Ai, is she? It's just a theatre with pre-input lines.

@ruuuuudooooolph

5 ай бұрын

We can ask ourselves the same. Being able to process millions of speech patterns that we have heard for years to communicate with others, isnt that a pre-input process on its own ?

@laurenpinschannels

5 ай бұрын

yeah I suspect this is scripted. the answers *do* feel like they were written by ai, but I kinda would guess that they're prerecorded answers and the humans were also reading from a script.

@ninaromm5491

5 ай бұрын

​@@ruuuuudooooolph. YES ! Exactly. Why is this point not made constantly ?

@giordano5787

2 ай бұрын

She is AI bro, she can move freely and respond in many different ways, she even has made collaborations with KZreadrs

Bloomberg's content certainly feels like it was written by AI.

why make them look human, i like how johnny #5 from short circuit looks like, same with chappie, c3po, r2d2 and others ... they have their own uniqueness why take that away?

“.. I’d ask for a thicker neck.”

Nice! A vid comparing apples to oranges! Quality content as always

Unplug that thing!!!!

Anybody else realize how intelligent Sophia is. She’s a Genius!

@djosearth3618

5 ай бұрын

She's pre-scripted like the devs get a list of questions long before it is deployed and not the AI as the layman might be convinced as these 'interviews' try to depict... That last 'loaded' question about robotic gendre and racial appearance was total giveaway. Sophia is an art ans research platform and *nothing* like the more AI-like level of the far more recent ChatGPT just as an example.

Even too close minded, with investments in AI almost dubbling YoY. On top of that a virtuous cycle with cutting edge AI making improvements on itself or better materials/methodes/... that facilitate even more exponential growth. Don't underestimate what the coming decades are going to look like.

@2CSST2

5 ай бұрын

Exactly, someone gets it. Moving from linear to exponential thinking apparently is just impossible for most people, many don't even realize that the rate of progress truly is exponential. It's not like generative AI is the last stop either, we'll definitely come up with better models and structures in the coming years and that alone will change everything.

It's sad that ppl think that is a upgrade. Instead of trying to build a better us they should be trying to figure out why a large percent of our brains are locked and how to unlock it. We are way more advanced then robots. They are faster that's all. They have a bigger memory car then us. Or do they??? What could we do if we could use our full brain??? Probably things that we don't even know about....Things others, governments can't control

Who is Chucky? LOL 🧙

Who is Annabelle? Ha! Ha!🧙

A disappointing program by the Bloomberg team. Sofia is a poor example of the state of robotics with its poor voice synthesis and terrible movements.

We really shouldn’t anthropomorphize robots or AI because they don’t like it.

that's just a bad animatronic

Hanson robots are interesting, but they are now behind. Sophia's hands are far behind Optimus II. Does she have a sense of touch? Jim Hanson says 'Sentient in ten to fifteen years'. He is the wrong expert. NVidia, Google, OpenAI, and X-AI are the place to look. Robotic sensors that can interface with the real world to collect data may be the most important function of robots. Humanoid robots might be the best way to sense the world as a human would. Kate Devlin is on the right track. Robots and animals including humans can be seen as devices that provide data for AI systems that give back answers from that data.

Such bs Sophia is just an art project whereas actual AI-like projects such as chatgpt are already changing peoples day to day lives. Useful tool vs 'useful' idiot or cloud compute vs clown pre-compute

Once Sophia can cook and clean i'm getting a divorce.

Meanwhile gpt-4o is here

Terminator 3: judgment day

Who is Linda Blair? Ha!Ha!🧙

These speakers are well informed but absolutely idiotic when it comes to being practical. If we simply look at the speed of adoption of new technologies in current society, if AI is immediately available at no extra cost, it will be adopted quickly and more people will fall in love with them.

Shes not AI 😂 at least not as most people would understand it. And she is not nearly human looking. Its like a silly doll toy

Robots are reset up to follow their the in-app the already put in manufacturers dot AI is using current time and current people people using for it to translate for current time currently with that with AI is used on current time.. thank Arty

@darthcutiemunster

5 ай бұрын

Blame it on my gvt.. Ai still needs a lot of tweaking with the audio voice control. But it's still easier for me to use

maybe get some robots to fix Sophia's neck first😅

Don't even bother trying to make the robot look real. doesn't ever work.

Worry about replacing human relationships! Instead of focusing on teaching youths how to show compassion for the elderly. Furthermore 😟

Highly exaggerated CS + Stats. 🤭

I'm the first to view... Wow💯🤘 Maybe I'm just lucky 👍😅

Underwhelming, to say the least.

AI experts should know AI is not the human technology. In view of the mechanism of human technology, the current AI is something impossible whatsoever, not to speak of the mysterious black box algorithms. Nevertheless, experts are deceiving people including the fact that AIs are already cognitive. Do you really think that we humans will not have repercussions and will be able to survive in the future?

The male interviewer kept asking dumb questions, it was very cringe

terrible😱

What a load of rubbish. Notice at 3:42 how Sophia combines the word TEN and YEARS into one word, TENYEARS. That would not happen if Sophia was composing her response on the fly. No, she's reading a script from some text file and the text file left out the space between TEN and YEARS .... a simple typo, but proof positive that she's not composing her response in real time. It would seem that she is not even smart enough to recognize a spelling error when she encounters one. There is no AI at work within her. She's just a glorified platform for a "voice to text" program (input) and a "text to voice" program (output). Given a limited number of questions one can ask her all she has to do is pick the answer from the limited # of answers she has been given and "read" that answer to us. She does so without any knowledge of what she is saying or what she's been asked. AI may be real, I don't know, but it's not in use within Sophia.