How AI Will Step Off the Screen and into the Real World | Daniela Rus | TED

[@cscs88]

We are literally on the edge my friends. It's going to be a wild , exponentially progressive ride. Lets hope for the best.

[@ammini999]

It would be better to make a decisive leap into the AGI to help those people in need.

[@mk1st]

“AI that learns after deployment” 😳

[@dekev7503]

Don’t believe the hype. Most of the “progress” that is being reported is marketing and the real practically applicable projects are few and far between.

[@businessmanager7670]

​@@dekev7503wrong. while there is hype, there is plenty of scientific evidence showing the advanced capabilities of ai and how far it can go in a short while. this is not like anything we've seen before

[@jameskim4028]

🤞

[@SinMihai]

Daniela Rus, romanian scientist genius woman. That is just brilliant, all the rest of AI's are overloaded with data and can't focus on the important task. Love it.

[@BestBuysReviewed]

Alien technology inventor

[@SecretEyeSpot]

The prospects of this new learning Model was understated in her presentation. What it should have explained is how it contrasts from Neural Nets, and how it can map itself in an allocentric fashion to recognize previously learned objects that it may encounter in an unforeseen contexts. Without a viewer's understanding of the current limitations it's literally unfathomable to understand what the Liquid Model intends to be without finding another video about it

[@XxZeldaxXXxLinkxX]

She went over the demo with locating an object in differently environments though. Either way, this is a short Ted talk and cannot be faulted for not covering every detail

[@wyqtor]

We Romanians are very proud of Daniela Rus!

[@californiaplant-basedeater2761]

How are you going to entrust the future of our species to anyone that doesn't look happy? She had the conviction of someone reading from a script. English isn't her first language but still, it was bad.

[@ovylord]

Lol. Lets be real. No one besides a few people knows about her in Romania.

[@ochiorbus]

I think only people who helped her to achieve her goals can say that.

[@Andrea-u2f]

Revolutionary Work. TY D.R.

[@Andrea-u2f]

Revolutionary Visionary Daniela Rus I thank you.

[@importantname]

please develop an intelligent machine that does my dish washing, clothes washing and folds it, cooks for me, puts the bins out, cleans the floors. So that i have the free time to do art: paint, write poetry, study and be at one with nature.... Not the current way that i do all the work, and AI gets to do all the fun things.

[@_Daio_]

There's a wife joke In there somewhere, but I'm too scared to look for It.😉

[@thinkabout602]

😂@@_Daio_

[@محمدالبيه-ي8ق]

انت ذكيه نعم الآله عبد و خادم وليس الإنسان كثر الجدل بسبب يريدوا احتكار الذكاء الاصطناعي لا أكثر عندما فشلوا في الكونجرس بدؤا يستخدموا أدوات و أساليب المخابرات الاعلاميه و التظليل ويستغلوا الحربين غزه و اوكرانيا

[@محمدالبيه-ي8ق]

You are smart Yes God is a slave and servant Not human There is a lot of controversy because they want to monopolize artificial intelligence and nothing more When they failed in Congress They began to use the tools and methods of media intelligence and shadowing They exploit the wars in Gaza and Ukraine

[@nickreffner4574]

@importantname This! All this talk about robots doing all the jobs and living in abundance. How about we start with the simple chores? Yeah having to go to work can be a drain, but it’s the house chores and maintenance that really drain you. Those 6-8 hours after work could be used on family, hobbies, or education all while robo-maid 1.0 keeps things in order at home. You could take the kids out and send a text to your robot telling them to have dinner ready in an hour. This would revolutionize the human experience all on its own. Regardless of breakthroughs in economics science and health.

[@NickDrinksWater]

The more ai transitions from digital form to physical, is gonna be the most impressive, and spooky. I Robot vibes

[@MaciejNowicki1989]

My wife Eve is perfect

[@netscrooge]

Rus is comparing apples to oranges, taking some cheap shots at systems like ChatGPT. Her system isn't better; it's better for specific tasks. If a robot is taking a philosophy class, having a more efficient understanding of the physical world won't be much help. But maybe this approach can inspire improvements in natural language processing. Who knows.

[@hyozanhades09]

Irobot combined with ghost in the shell with the new cybernetics coming out from nuralink

[@FSboy70]

Maybe you get a n offer to be uploaded to the digital world so you can live forever and just discard your body all together?

[@AppleOno]

The most important part of this isnt getting these machines to work, it's securing them against being used to cause harm.

[@marcsequence]

@stephanieellison7834 word yo

[@alexxx4434]

In the current economy paradigm AI will play a role of replacement, competition for humans, not as much as helpers.

[@muteloch2798]

This. Would we expect charity from corporations? Unless the robot is a public service we can expect the helpers will (first) be reserved for the super rich and ultra wealthy.

[@raoultesla2292]

There is no competition. Robot no drink, no sleep, no distractions of any kind from goal.

[@alexxx4434]

@@raoultesla2292 And no salary.

[@LAWalster]

Extraordinary and equally terrifying…🤯

[@fernsehdesign]

Russia likes all this, too… 😳

[@SecretEyeSpot]

What about this terrifies you?

[@fernsehdesign]

@@SecretEyeSpot what could this be… hm… can t think of anything

[@muteloch2798]

My mind is blown as well. I'm reevaluating my career priorities.

[@jammaschan]

@@SecretEyeSpot the potential for misuse

[@boremir3956]

How scalable is this? Can they go from a few neurons to a large network? Because what she has shown so far are very small models. Also how do they compare to standard neural networks in terms of energy efficiency. Scale and energy efficiency are two important things she has not mentioned.

[@Johanneslol11]

Openai is busy with building chips that are specifically for this. So that will stop the biggest issues you mention here

[@boremir3956]

@@Johanneslol11 She mentioned her team is doing something different from the LLM's such as those from OpenAI.

[@MikeHodgkinson]

gpt4 thinks: Daniela Rus's TED talk on liquid neural networks introduces a groundbreaking concept where neural networks are not fixed in hardware but exist in a fluid medium, allowing dynamic reconfiguration. However, in her presentation, the scalability of these networks and their energy efficiency compared to traditional, fixed neural networks were not directly addressed. Scalability: Liquid neural networks, being in the nascent stages of development, primarily demonstrate feasibility with smaller models. The transition from a small network of neurons to a large-scale network poses significant challenges. These include maintaining stability and reliability of the network connections within a fluid environment and managing the complexity that comes with larger network sizes. Energy Efficiency: Liquid neural networks could potentially offer improvements in energy efficiency over traditional neural networks. Traditional networks often require a substantial amount of power for operation and cooling, especially as they scale up. In contrast, liquid neural networks, depending on their physical and chemical properties, might utilize less energy due to their adaptive and potentially simpler cooling needs. However, concrete data on their energy consumption compared to solid-state networks is yet to be clearly established.

[@FusionC6]

@@MikeHodgkinson yeah but its not openai so people wont care unfortunately

[@necessaryevil455]

Human beings don't respect the physical autonomy of each other, robots are going to have a hard time.

[@NamrataLamichhane-fv6dg]

Very true

[@diybeast3943]

I’m sure the majority of people will understand that robots arent as advanced yet to think exactly the way a human can. More understanding less anger

[@XxZeldaxXXxLinkxX]

Don't anthropomorphize machines...

[@BrandonMcCurry999]

​@@XxZeldaxXXxLinkxXyou do realize that ingl group/out group bias is the main cause behind many... many wars ect throughout human history Don't you?

[@BrandonMcCurry999]

​@@XxZeldaxXXxLinkxXI thought about this occurring, the possibility A world ending one, I "solved" it with synthesis

[@GamerBoyRobby]

Large scale end-to-end liquid network training is the next big thing it seems

[@machinegod3000]

Cutting down the demanded costs of AI progression makes it even more faster - humans are too eager to outdating themselves! Imagine, within this decade an AI or a Robot gets granted personhood. That means alot!

[@bloodust7356]

I think the robot named Sophia already have been granted citizenship in the UAE.

[@machinegod3000]

@@bloodust7356 yeah, its symbolic...but there is a difference when you are also a recognized individuum with all rights and duties.

[@Tijaxtolan]

I think only westerners and particularly some Americans think this way, overly materialistic, they value the results more than the process, those people would sell their family for robots if tech companies told them it's "the future"

[@SuperStargazer666]

Things are finally getting interesting. Bring it on!!!! Full speed ahead.

[@TusharKale9]

Amazing to learn what we can learn from such small things in life too. Thank you for sharing

[@ilytyspirituality]

Skynet approves this message

[@TheYashakami]

Making something that doesn't make mistakes is a fool's errand. You should be making something that can respond to its own mistakes.

[@Evfaint2089]

The point of ai is to learn from itself essentially the same thing

[@GaryMillyz]

yeah you tell 'em, einstein 🙄

[@GY-ez2ww]

It's amazing achievements! AI is moving faster than people imagined. But I think AI for public benefit needs more attention.

[@godmisfortunatechild]

Remember if you loose your job and become economically superfluous, due to AGI, there's no law guaranteeing you'll have basic necessaties met, as things stand now.

[@gridvid]

Impressive. Liquid Networks are the future

[@duncanmaclennan9624]

EPIC. So good to see her on stage again

[@timgiraudier6399]

The military is going to love this, combat models coming soon!

[@BudStickly]

No way possible that the military doesn't already have it.

[@andycordy5190]

Physical intelligence! Wow! What a potential!

[@favesongslist]

Also know as "embodied AI".

[@renvictor]

When I saw the image of an AI robot assisting on pedestrian lanes. I immediately assumed that people would just run over it, especially in our country.

[@heartyfisher]

LOL they will have cameras and record the event in HDTV !!

[@Wigglewonks]

MIT’s biggest grant will come from the department of defense.

[@Vpg001]

I didn’t watch the full video yet (35 seconds in), but I think neuralink is a key way to teach robots how to act human. At the same time, data collection is overpowered and needs to be regulated!

[@dreamphoenix]

Amazing. Thank you.

[@TheTuubster]

BTW if you are a Star Trek fan, you already know Generative AI since the late 80s. The Holodeck is EnvironmentGPT, the Replicator is FoodGPT, the compute is ChatGPT (with a voice interface) and the Universal Translator is LanguageGPT (training the voice and language of your partner, noise cancelling out the original voice and replacing it with the voice transformed to your languange). If you want to know what Generative AI is going to do for you, watch "Star Trek".

[@TheTuubster]

@stephanieellison7834 The text raises several speculative and unsubstantiated claims: - It provides no credible evidence that governments or corporations have a secret plans to deliberately cause mass starvation or "extinction" of the population through lack of jobs or income. Major policy decisions are usually public and gradual, not secretive or catastrophic. - The assertions about declines in farming production or changes in robot usage away from the public are not backed up by any data or sources. Broad societal changes usually have multiple interconnected causes, not secretive plots. - Questioning the rationale for farmland purchase in Maine on the basis of chemical levels far below safety thresholds, with no other context or evidence, amounts to an unsupported conspiracy theory. - Predicting violent reactions from other countries based on this hypothetical scenario is highly speculative fearmongering without factual basis. - Overall the text presents a dystopian population control narrative as proven fact without credible evidence. Conspiracy theories typically ignore alternative explanations and lack consideration of societal complexities. In summary, while job disruption from AI is a serious issue warranting discussion, the text's claims appear to be unfounded speculation rather than supported analysis. A more measured, evidence-based approach would be needed to evaluate impacts and policy solutions regarding AI and employment on such a large societal scale. The tone and lack of sources undermine its credibility.

[@TheTuubster]

@stephanieellison7834 Here are some of the logical fallacies contained in the text: Conspiracy theory - It asserts secret plans by governments/corporations without evidence, ignores simpler explanations, and lacks consideration of challenges of coordinating such a vast conspiracy. Slippery slope - Suggests modest actions like farmland purchase will inevitably lead to the totalitarian control scenario depicted. No evidence more gradual changes are impossible. Hasty generalization - Questions one state's land purchase to make broad claims about national policy and societal changes without adequate sampling or analysis. False causality - Implies job losses from AI will directly cause starvation without jobs/incomes, rather than multiple social factors influencing employment, welfare, technology adaptation etc. Strawman - Portrays opponents' views/policies pejoratively as intentionally designed for "passive killing" or "extinction" rather than more nuanced policy debates. Undistributed middle - Assumes specific impacts like starvation and high job losses from one cause (AI) without ruling out other possible contributing factors. Appeal to fear - Uses speculative scenarios of violence and social unrest to argue its point rather than substantive evidence or reasoned policy analysis. Anecdotal evidence - Relies on isolated examples like the land purchase rather than comprehensive data to make sweeping conclusions. So in several ways, the arguments violate principles of logical reasoning through fallacious assumptions, causal oversimplification, lack of evidence, and manipulation of emotions. This undermines the credibility and objective evaluation of its dire claims.

[@agneevdey]

did you see that? "1:36" keep your focus on the left hand side of your screen below the ted logo. notice a weird shadow pass by. "1:41" the man in specs seated right next to the logo even notices the thing or maybe sees something from the corner of his eye as his reaction is a bit late but he does react.

[@Leto85]

That makes sense to develop an AI that teaches robots how to operate tasks instead of us trying to teach them. AI is much closer to tobotics than we are ourselves.

[@Leto85]

@stephanieellison7834 I think money itself is the problem. Once AI takes over and do these jobs we have no need for money anymore. It'll be better for us when things can be provided for all. But that would only work if we could teach AI some empathy. Without empathy there's not much to live for anyway as things won't matter then. And I am interested in what goals AI would have with its high intellect.

[@MythicalBeserkers]

Loved it. Really interested in these

[@nextsunnday]

The liquid neural netwok looked interesting, but there is always be slight delay in the reaction speed of robots when interacting with real world and i think robots with onboard memory and intellgence is not yet ready as most of the robots rely on huge data servers to make decisions for them.

[@The_Unexplainer]

Hardware issue, solvable in theory.

[@Ghost-pb4ts]

It is possible to solve complex problems using machines that can perform faster than any bio-organic being in the world. The speed of processing information by photons in your brain is limited only by the constraints of biology and environment, not by physics. However, machines in the future will have limitations based on physics.

[@aussiepawsborne9056]

Tesla fsd runs locally and it drives vehicles in real time. The computer is the size of a PlayStation. No reason why a robot couldn’t pull inference locally and fast speeds

[@philipehusani]

Sorry, you're wrong.... A robot is hardware, you can deploy a model on it and give it memory just like any computer, if I can use LLMs on my laptop, pretty sure robots will have no issues.

[@e.v.k.3632]

Don't forget the risks That's why we have to get better too with gene editing

[@NeorecnamorceN]

Or not. The risks of gene editing are an order of magnitude greater. Maybe let's focus on actually understanding our body and genes before we leapfrog to editing them?

[@e.v.k.3632]

@@NeorecnamorceN We know already a lot and we improve plants and animals for a long time There is also gene editing what isn't transferable to the children So there are no risks You're just scared Also gene editing is already common to cure humans So why not finally improve us

[@NeorecnamorceN]

@@e.v.k.3632 There's a lot wrong there that I just don't have the energy or interest to get into so I'll just say this, there is an order of magnitude difference between modifying a single problematic gene and modifying enough of our genetic code to exact any kind of specific change to our physiological system in part or in whole.

[@TheTheCriticX]

'Liquid Network'- New Type of adaptable Ai

[@TC-en7vi]

The only stone in the middle for reaching this technology for all not for a few is money. The only real problem that separate us from greatness.

[@russelleverson9915]

I'm not sure how to feel about someone speaking with so much optimism and excitement on the topics of: 1. Creating a type of AI that continues to learn autonomously, and 2. Putting that AI into physical machines. I'm wondering if someone should sit down and watch the Terminator movies with her. 🤔

[@Pearlylove]

Good point. This scifi movies were created with help from excellent scientists, so both robots and future tech scenarios are VERY realistic. Same with other topics, like chimeras. Motives are always dual/several. One is to form your opinion/brain of what to accept. Today, sci-fi is reality, and what is presented from politicians and media is only the top layer. We all should dig a little deeper, and speak out with what is not ok, and there’s a lot. Live your life with your eyes open, so you can say to your kid that you did all you could for her/him to stay human, have a home and a job.

[@KrishnaSingh-rd6pr]

AI AS HELPERS >>> AI AS Replacement >>> AI as human killerz

[@merxellus1456]

I dont care what harm or danger ai robotic uprising might cause.. i just wanna see it happen cos its cool af

[@merxellus1456]

@stephanieellison7834 nah man.. im not reading all that

[@PedroHenriquePS00000]

the problem with AI today is that its text based only, meaning that it compares to those old computers that only ran ms:dos with green screens and walls of text... current ai is basically that. i think advancements will come when AI has the ability to BE a visual interface note that i said BE a visual interface, not HAVE one... because "having" a visual interface means that a human coded specific functions so that a user can use it, what im talking about here is the ability for ai to CREATE ui and ux on the go, as you request stuff and use it.. without the need to download anything, you shape your OS as you go. you can always have predertermined shapes, but im talking about a fluid os experience. what does that even look like? well think you a blank screen and a microphone, you say you want to create music, and a basic music software window appears, but as you use it more and more options show up and you can alter where buttons are and where components go... when youre done, you can save that music experience as an "app"

[@PedroHenriquePS00000]

hopefully this sparks someones ideas... we need more than mac, windows and linux, its time for ai to join the game of OS i also think this AI OS could be the core of service robots

[@sailingintosunshine]

current AI is absolutely not textbased only, models are trained and deployed on all kind of inputs, from text, images, and numerical numbers to visual interaction data, real time camera input, RNA, the list is endless. Anything you can convert, interpret, encode or tokenise into numbers eventually is usable.

[@DjHazardous]

*For now i would say but its learning quickly in some aspects like music creation, video creation, company agents and a few more that text base will be phased out within 10 years then you have tiny hints that there's much more advanced AIs but are currently being withheld from the public as they pose a significant danger to our current economy or latest security measures*

[@favesongslist]

Embodied AGI. Liquid neurons will enhance this to become reality.

[@humann001]

3:22 this ad for Mercedes is far from ideal.

[@chidieberemicah6322]

While Nvidia is working hard to create cheaper and lighter hardware for AI processing, Daniela Rus and team are working to lower the hardware and compute ceiling for using AI via "Liquid Networks." These 2 concepts intersect and the future is simply crazy🦾😲

[@TheTuubster]

Generative AI is already a large part of my every day life. I post process or transform my photos with it, I have corrected medical diagnosis with it (which my doctor agreed to) or solved a breach-of-contract conflict without using a lawyer. And if someone trolls me or others, I use it to generate an analysis of the ethical or logical flaws in their statements and let the AI deal with them (and it easily discloses all the flaws in a statement within seconds).

[@TheGreatDutin]

Yet it's not always right, there's a human somewhere deciding what it can and can't say. Politically correct AI isn't very useful if you ask me.

[@favesongslist]

@@TheGreatDutin The power is kept only for the large cooperation's that use the raw models.

[@TheTuubster]

@@TheGreatDutin That is true, since Generative AI is based on statistical models and filtering/altering the statistical parameters distorts the consensus represented in the generated result. It's like preventing a polling institute to make racial bias in their interview data transparent. Filtering Generative AI results is a form of censorship as statistical truths are kept hidden. The good thing is: The models of LLM can be run on a local computer with a modern graphics card since they are only a couple of gigabytes big and fit into the memory of such graphics card. You can get unfiltered chatbot results this way. But of course it is much much slower.

[@TheTuubster]

@@favesongslist You can use raw models on your local computer, "llamafile" is such software.

[@TheGreatDutin]

@@favesongslist Yup

[@ekanshsudha]

This is something new which I have seen on KZbin after such a long time. This motivated me.

[@gg5777]

Those machines and programs will soon start talking to each other not waiting for the releases. Allowing them to connect will create speed that we haven't seen before.

[@user-tc5pl3zw3h]

"...imagine being surrounded by helpful robots at the supermarket." OK, then imagine Skynet becomes self-aware, and those robots surrounding you decide to no longer be helpful. You're still surrounded by robots at the supermarket.

[@rlews1531]

This is friggin brilliant. It seemed obvious that brute force simply couldn't be the best we can do. Don't get me wrong, the reiterative trial and error (what's the getting hotter - getting colder childs game?), was a massive breakthrough. Showing something that most thought couldn't be done, actually could be done, in a big way, remains the pioneering work. Trinity and the manhattan project. But, there are some super, super smart people and once they "aha" moment of using physics for reinforcement learning - BOOM!!!!! We're rocking and rolling now my friends. Great thinking Ms Rus. Get some investment and pedal to the metal...!!

[@ajithboralugoda8906]

Brilliant and counter intuitive to crunching LLM , iron shod!

[@jaidenBenzz]

Great talk thank you

[@AdvantestInc]

Integrating AI with robotics to achieve physical intelligence could truly transform industries. What might be the first practical applications we see?

[@mohammadrahimjamshidi79]

AI, in order to improve its performance and prevent undesirable consequences, must continuously interact with “effective rules and stable principles in the realm of existence”. X Mohammad Rahim Jamshidi

[@wffj-tv2652]

I think I saw this movie, already. It was called "The Terminator."

[@mrpapafakis1290]

Honestly, as nieve and kinda silly as it sounds, I think we should only create AI robots with limited capabilities, like the ones that appeared in Star Wars. They're only useful at specific tasks, very beneficial to humans to have around, but have no aspirations of paperclip farming us into oblivion. Whereas, with the super AGI proposition, as it stands at this point in time, comes with too many potential downsides. They might be able to find cures to cancer and design FTL engines, but they could equally doom us to an Animatrix style future as well. Just not sure the latter is worth the risk.

[@Raulikien]

You won't put an AGI into a house robot but it'd still need some degree of generality to accomplish real life tasks

[@TheHippie27]

It's really sweet that they had those cute cartoons with robots helping grandma cross the road and giving some rock climber a power suit, but let's not forget that this is all going to be used by global military powers, and has already become an A.I. arms race. I guarantee she's given this same talk to all the big world leaders and their defence secretaries, and they're all imagining all the ways it can be used to give them an advantage over their enemies. This is gonna get so messy..

[@danvorosmarty9854]

They already have AIs that have been trained on huge numbers of simulated dogfights. They are better than any human pilot at piloting a jet fighter. What country wouldn't deploy that technology in war?

[@mka11221]

Keep it going TED .... AI is the hot cake

[@edgaddi7369]

That was fascinating .

[@shawnvandever3917]

This is great. However they have been saying this for a few years now without a big working demo. Continuous learning would be massive. Hopefully this works out

[@BEN_MONK]

Very brilliant dedicated programing approach on the way of GAI 👍

[@micro8106]

From this video I can received the moral how to make a work more easy. In the future it is necessary for us to mingle with the machines and AI. These products and function could reduce the tensity of the task. But we need to give them the right information and mandate. As speaker mentioned before important of the Liquid system. Not only that but also other system, for example like a weather or air we also can utilise this component to make a better machine.

[@TravisCotter]

One day you will see the Son of X coming in the clouds with great power and glory. Xman

[@NeorecnamorceN]

There is also a safety benefit to having AIs run locally. It prevents future AI viruses from infecting the entire Network.

[@AnnoyingMoose]

A.I. can currently turn text into images but when will it be able to describe an image using text?

[@joshuajohnson9914]

Welp boys, this is it. The end of an era and beginning of the AI. Before I was still skeptical but with this new stuff well I’m now convinced. I for one, welcome our future Ai overlords and wish them a comfortable bug free existence.

[@gonzalosanchez4572]

I feel like I'm on a rollercoaster...

[@myaubeats]

Ai that inproves itself without humands knowing what and what is changing sounds like a great idea. Millions of AI systems "randomly" adapting at all times, what could possibly go wrong

[@mikeroy9316]

Off the screen and on to reality, and the energy to achieve this fantasy will come from where.

[@emilianohermosilla3996]

As an American the only thing I have to say is “Y’all better be getting that meaning economy working ASAP”

[@rohitlayne]

Wow! This means the AI doesnt have to be expensive. I can get a basic robot, train it for a week or so about my house chores, and it will take care of it. Sooooooooooo good!

[@kusumayogi7956]

AMAZING!!!! ARTIFICIAL BRAIN BY MIMIC WORM'S NEURON THAT IS SO COOL

[@FuraficFark]

This is the scariest Tedtalk I've ever seen 😮

[@ProgressiveEconomicsSupporter]

Great achievements!😊🙏 But, please also regard that simple AI already seems to help autonavigate war drones in Ukraine war, so they are not easily disturbed. This can be great for defense purpose, but worse in the aggressors hands

[@gb8602]

I think AI shouldn't be used in wars, but I'm against war in general...

[@mr.murray7786]

Personally, I’m not a fan of AI. I believe it makes life, problem solving too easy. I enjoy being as independent from technology as possible. A free thinker. But I’m grateful it’s there to support those who need it. I just hope we don’t lose wholesome, human-to-human interaction in our ever evolving society.

[@Allofussurvived]

Your ideas don't matter only mine do

[@Allofussurvived]

We need AI For UBI that's it nothing else

[@imadeyoureadthis1]

A.I. will surpass us so you'll get to have much more wholesome interactions with it. Humans have many many flaws and limited abilities to control them. It will create a better moral system than the one we have now. Just look at all the suffering and pain we cause each other, mainly caused by our immaturity.

[@love.connection_1]

​@@AllofussurvivedUBI MEAN?

[@Brian-oz8io]

I like to think of it as.. it helps us with basic problems so we can focus our attention on bigger things, even more difficult challenges and problems that if solved, will result in huge net benefits to humanity

[@shermanwellons]

We are celebrating the birth of AI, and the beginning of the end of our existence as the human race on this planet. Wow, someone needs to rewatch The Matrix!!!

[@nunyabiz9952]

This is creepy, I remember teaching myself how to drive I would focus on the horizon lines too. This a.i. is teaching itself the same way I did. Wow! Are we talking down on how it drives? Its teaching its self! And I've never crashed before and been on the road over 20 years. Imagine how good a.i drivers will be❤

[@markseagraves5486]

When we as humans choose to freely embody the basic moral and emotional intelligence to recognizes the rarity and value of all life, perhaps we can teach these truths of Insight, wisdom, loving kindness and compassion to the robots as well. Perhaps they will listen, learn and act better than we have so far.

[@edh2246]

The best thing a sentient AI could do for humanity is to prevent us from killing each other, not by force, but by disrupting supply chains, communications, and financial transactions that enable the military machines throughout the world.

[@seanLeprechaun]

Are the same people who are wowed by what can go right with this also considering what could go wrong?

[@RobertAlexx]

Bravo. I count on robots.

[@juliusapriadi]

TED successfully produces lifeless presentations by enforcing its now standardized speech structure onto the most interesting people. Instead of this stiff approach, can TED maybe move towards a liquid neuron approach with those talks, to allow for speeches that are a better individual fit for topic and talker?

[@AIHahb]

AI is changing our lives.

[@zeekjackson2483]

This comment is for those who have theories, speculations, and ideas of how this will end the world as we know it... will it be like terminator? How long till it's too late? Can it be stopped if so ? Is it possible to learn how to control it indefinitely?

[@Intellectualodysseyai]

We are only a few years away from the most significant transition in humsn history and life as we know it. Let us manifest a peaceful word of abundance. We are going to witness a life with universal income and a purpose that includes choice, not based on the availability of workforce. I cant wait to tell the new generation what life was pre-transition

[@ConorRyan-kw4sx]

Gosh, I just can't wait until we apply this technology to autonomous weaponry (Soon?). What could go wrong?

[@cujo1050]

The potential for calamity when AI moves into the physical world is extrordinary. I do hope there is video out there showing Daniella highlighting the dangers of AI and what rules and regulation need to put into place to promote change in a way that significantly reduces the risk of humanity revolting against the technology or the technology deciding mankind is no longer necessary. It is unclear to me that she has realization that the speed at which people will be displaced in the workforce will be unprecedented let alone the amount. Also unclear is if she understands how those people will react to this and the civil unrest they could cause. That course of action is every bit as likely as the possibility of retraining and in my view the former is more likely that the latter.

[@dorpersatdawn]

I don't think this lady has seen Terminator...

[@Slayr.]

This liquid networks thing might be another breakthrough.

[@j.d.4697]

1 step closer to replicants. I can't wait!

[@psikeyhackr6914]

I am terrified of any future of computing that would come out of a building that looks like that. 1:06

[@ChilsonTV]

“All normal life, Peter, consciously or otherwise, resent domination. If the domination is by an inferior, or by a supposed inferior, the resentment becomes stronger.” ― Isaac Asimov, I, Robot

[@matthiaskatze1932]

We have to take measures that AI does Not take over the world. Neverthrless I research how AI can get all means to do so in order to pay my rent and expensive car. And to comfort my guilty conscience I do TED presentations. Makes Sense. From an individual perspective.

[@deeplearningpartnership]

Great talk.

[@hansolav5924]

Before we give AI the slightest chance to take a look around and make judgements on what they see, we should make sure to first become a species worthy of life. We are not, at present. haven't been for some time. Ever since we learned how we are strangling our living requirements on this planet...and did not even slow it down.

[@sigh1685]

We're near the singularity bois and gals

[@justsayingpod]

Why do I feel this will be the moment we look back on and say this is when it all changed haha!

[@alexj9111]

Conventional computers are already beating grandmasters at chess, just imagine what a robot controlled by a quantum computer could achieve. I hope i'm still around to witness one making a 147 break at snooker.

[@da7me22]

Quantum computers work differently. You won’t see an AI built by quantum computers any time soon.

[@alexkava9318]

Imagine robots in super market asking you for “tip” after helping you! Thats sound more realistic for the future.

[@DavidGalich77]

My thoughts fall along the lines of humans growing along with AI through the physical world. I wonder what kind of wondrous places it will take us.

[@Forshledian]

The technology that we (included our AI tools) will develop in the next 100 years is going to be insane. But what is the world going to look like in 1000 years from now? What about 10,000?

[@andra9694]

Let’s hope the world survive this coming 10 years first

[@s1v7]

Bravery in our days is the most valuable trait.

[@kylev.8248]

There is a very thin line between “Bravery” and ignorance

[@BenjaminMora-r8c]

Can someone please come up with differentiation between algorithms and AI? Because as best as I can tell, AI was invented over 2,000 years ago and it continues to be absolutely harmless and not sentient at all.

[@jsacodes916]

A world of radical abundance = a world over run with waste.