10 disturbing AI breakthroughs

So today we’re going to take a look at  the year 2024. in AI space everybody  knows just how crazy the year 2023 was  and now we really do need to look  forward to what we could potentially be  seeing in the future I think this  ensures that we aren’t caught off guard  by any new AI Technologies or Trends so  let’s get straight into this one of the  first things I think that we’re going to  see is the rise of AI influencers and  verification techniques for instance  there is a Spanish influencer created  entirely by AI that generates its  modeling agency 9 ,000 a month with  200,000 followers on Instagram now it  might not seem to be an issue on the  consumer side but if you’re a Creator on  Instagram and you’re spending so much  time creating content for your followers.

  it might get a bit frustrating for you  when you see these fake AI influencers  getting so much engagement like this  fake Spanish influencer it also means  that in the future not only are there  going to be AI influencers trying to  earn money but I do think that we are  going to be seeing something where a  large majority of the Internet isn’t  going to be real, for example if I did  see see this person on the internet I  would assume that with 200,000 followers  they are real because of the rise of AI  influencers where people have around  200,000 followers and it isn’t disclosed  I do believe that this is going to  fundamentally change the internet  because right now on Instagram Twitter  and YouTube you could be an AI  influencer without needing to disclose  anything at all there are no legal  requirements no European laws and no  digital laws that say you do so.

however  I do think that in the future this is  going to change this this is really  important because if you remember  several months ago an image of Pope  surfaced on Instagram which everybody  thought was real but it was an AI image  right now there were so many things  floating around on Instagram that we  don’t know what’s real and what’s fake  this is a real problem because for the  average person like you and me who’s not  in the media Spotlight we don’t have an  issue because our faces aren’t plastered  all over internet however for people who  are famous they continually get altered  and their likenesses are used in these  AI systems so it’s a real issue you when  these things happen now you might be  thinking okay they’re famous they should  have signed up for that that’s what they  signed up for but think about this  imagine you’re a normal person with a  small account and then someone starts to  use that technology to make you do  something you haven’t done and say  things you haven’t said your face might  be used to promote things which are  illegal and it might get you into a lot  of trouble.

so this is going to affect a  lot of regular people and this is going  to be a real real issue take a look at  this this is a new Aid driven Avatar  animation and you can see just how  realistic this looks and remember that  this technology is only getting better  and the difference between real and fake  is fading real fast now take a look at  this so here we have this AI product  where essentially you just need a single  image and a driving pose and you can  essentially make someone dance or make  them do pretty much anything so in the  future it is going to be difficult for  social media platforms to verify a  person’s identity because a simple AI  program would be able to fake anyone’s  identity Now we move on to the next big  thing we might get to see in 2024 and  that is realtime text to speech this  specific AI will lead to lots of scams  in 2024 but we’ll come back to that  later now the problem with current text  to speech is that a lot of the times it  is limited in terms of how it sounds and  how fast it is but these systems are  getting faster the apis are getting  better and the latency is getting  shorter and shorter so what we’re about  to see here is play HT 2.0 Turbo which  is the fastest generative AI text to  speech API.

this essentially means that  if someone back engineered this we could  have real-time scams being absolutely  crazy you’re about to see how this is  absolutely insane I’m going to ask you  guys to listen to it but don’t focus on  how the quality of the voice sounds  although it’s good what you need to  listen to is of course the latency in  which it is played and the latency  essentially just means how fast it  responds to your message so let’s take a  listen to this hi there I’m calling in  regards to the purchase you made last  week hello Play support speaking  please hold on a sec and let me just um  pull up your details real quick can you  tell me your account email or your phone  number what you just heard there just  goes to show you how crazy this is we  got like 100 milliseconds response time  in a simple message so if you combine  that with a model that’s able to  immediately read your message and is  trained on thousands of persuasive scams  to convince someone that’s also trained  on a voice of one of your relatives that  could be a serious serious issue the  problem is remember you only need a  couple of seconds of someone’s voice to  train a model that is exactly like yours  some softwares can literally take 3  seconds of someone’s voice and make them  say absolutely anything.

so with that  being said this is going to be a real  issue in the future and I’m not sure how  we’re going to combat this just yet now  future scammers are going to be using AI  Technologies and it’s going to be really  really convincing this is a huge problem  because when you get an email you easily  ignore it thinking okay that’s a dodgy  email but if you have a relative that  messages you or is talking to you in  real time on the phone and is saying you  need to do this you need to do that how  are you going to verify that so 2024 is  going to be a year full of AI scams you  better watch out for that now let’s talk  about the next important thing and that  is humanoid robots with Tesla entering  the robot industry humanoid robots will  be the biggest highlight of 2024 take a  look at this gr1 robot this is the  unitri humanoid robot and according to  some tech platforms it is going to cost  you around $80,000  like you’re seeing right here he  literally kicked the robot and it’s  still walking it is able to balance  itself and that is a truly amazing feet  now you’re seeing he’s giving it some  real good kicks from the back too and  it’s able to balance itself without  falling apart what we’ve achieved in  last 2 to 3 years in robot industry is  crazy now imagine what we’re going to  achieve.

in the next 10 years and what’s  even crazier is that we’ve been watching  all kinds of robots from Last 5 Years  but they were all super expensive but  now with all these new companies  entering the market the robot prices are  coming down really fast now what’s even  more interesting is that these robots  will be combined with large language  models like GPT and of course this is  going to lead to some real-time  conversations with these robots which is  going to be really incredible and  essentially in 2024 what’s probably  going to happen is we’re going to have  advanced robots moving at an accelerated  Pace because we have a platform called  Nvidia Omniverse where you can basically  get synthetic data through millions and  millions of  and then you can use that synthetic data  to train your own robot using Nvidia  Omniverse they taught a robot hand how  to spin a pencil which is definitely a  very complex task for a robot hand and  this is the result they got after  hundreds of thousands of attempts Now we  move on to the next big thing coming out  in 2024 and that is the first ever video  game NPCs.

in 2024 we will see AI powered  characters in games with generative AI  these characters would have the ability  to self-train and self-evolve so  essentially they’ll have humanlike  intelligence now I really don’t know how  this is going to go down but it’s going  to be one of the most insane things ever  here what you’re about to see is a demo  of the first video game  NPCs yes what do you want do you know  what an NPC is no I don’t should I you  should you are an NPC what do you mean I  don’t understand an NPC is a  non-playable character you’re not real  not real what are you talking about this  is all so confusing what are you trying  to do man I’m trying to figure out why I  am here and what I should do next uh so  why are you here I’m not sure but I have  a feeling that I’m here to fulfill some  kind of Destiny Okay uh do you think  that we live in some sort of simulation  maybe I just don’t know for sure yet do  you think that uh you real I’m real I’m  not sure what to believe anymore what if  I told you you’re in The Matrix buddy  that would explain a lot but I’m not  sure I want to believe it what if I told  you you’re an AI that would explain a  lot but why would I be here maybe  because you’re programmed to be here  that would make sense but why was I  programmed this way okay so my story is  basically that you’re an AI

oh what do  you mean by  that you’re an NPC in a video game that  sounds frightening is that the truth  it’s the sad truth I see this is a lot  to take in are there any other answers I  should  know um not really this is very  overwhelming I’m not sure what to think  or feel right now yeah I mean you’re not  really real that’s hard to process does  this mean I don’t exist um you exist  sort of what does that mean am I real or  not it means that you know after I close  this game you don’t exist oh I see so  I’m only here for as long as the game is  running that’s difficult to  comprehend yeah and then once you walk  past this border you’re gone gone that  doesn’t sound promising what does this  mean for me now that was something that  I found to be completely just the most  fascinating things I’ve ever seen  because these AIS were having an  existential crisis about not being real  so yeah it was definitely fascinating to  see this video because these AIS were  having a complete meltdown knowing that  they weren’t real these games are like  creating another world where characters  have the same intelligence as humans with AI  entering video games we might get to see  some characters doing really crazy and  unethical stuff maybe that’s the reason  why steam has banned all AI video games  from its platform so with that being  said I do think that someone out there  is going to create the first video game.

I don’t think it’s going to be on Steam  I don’t know what platform it’s going to  be on it’s definitely not going to be on  Xbox or Playstation but I do think that  it’s going to be like a solo kind of  developer where you can play a simulated  game but either way I think that  probably will come and I think what will  happen as well in terms of The Wider AI  industry in terms of gaming is that  maybe one person is going to do it and  then monetize it and then that could  spark the whole entire gaming Revolution  but it will be also be interesting to  see how companies deal with AI entering  the gaming industry because as you know  if an AI does something that’s  unpredictable or something that the  companies can’t predict  there’s no way to ensure any kind of  safety so if they get the AI to do  something that’s you know completely out  of character.

it might invoke some kind  of legal trouble for example this video  game no man’s Sky which was supposed to  be a completely generative game it was  basically a game that was generated  generatively in a sense that every time  you explored an island it was supposed  to be completely new and everything was  supposed to be new for the user and it  was supposed to be impossible for you to  get a duplicate world but I think it was  just a little bit before its time but  now with generative technology coming  into effect this game is definitely  going to move in that direction you know  combining it with the AI generative  capabilities with open World Generation  Now we move to the next major thing of  2024 and that’s of course autonomous  agents so this is a clip from Sam Altman  and he’s actually talking about  autonomous agents and stuff like that  here we have a clip of the MIT director .

Daniela Russ and open AI CEO Sam Altman  discussing the future of autonomous  agents  so let’s first take a look at this clip  language interfaces that can control  agents is certainly another logical step  that people really want um the answer  will be like start slow and you know set  the training wheels very low and then as  we understand the risk surface and where  it’s okay and where it’s not we can make  thoughtful risk adjusted decisions about  how to move those training wheels up but  the upside here is going to be  tremendous for people um if you think  about having an agent that you can give  a task to to help you with whatever  you’re trying to do and be confident  it’ll go do it uh that’s awesome Daniela  I really like the idea of Agents um  whether embodied or not that can help us  with with cognitive and physical work uh  the embodied agents would help us with  physical work and the non-embodied  Agents would be like our uh our super  assistants and so the question is how do  we get to the point where these agents  can actually deliver on the tasks that  we have to do and so in the future I  really imagine progress in in connecting  the tools that we have right now with  more extensive capabilities uh that will  uh give the agents the power to do  better interaction with the physical  world that will uh give the agents

uh  the ability to do common sense reasoning  and to do um human understanding perhaps  at some point we will even get to AG who  can understand us um emotionally but  there is a long way to go before we get  there so that was the video now these  autonomous agents are going to be really  hard to do because the problem with  agents is that not only are they  autonomous but they also need to make  the right decisions consistently over a  long period of time if the agent commits  a mistake along the journey the computer  could freeze and the whole system could  fail and that could happen could stop an  autonomous agent from being successful  so it’s going to be interesting to see  how aut agents do work in the future and  how we manage to tackle that problem  because what we do know is that several  Tech Giants are spending lots of  resources to somehow achieve this feat  now take a look at this demo so this is  a self-operating computer being  controlled by an autonomous agent so  agent is asked to go to Google Docs and  write a poem about open-source software  and then you can see it going ahead  browsing google.com then it opens Google  Docs and starts writing the poem so  essentially this is what I’m trying to  explain that a autonomous agents are  going to change the internet  fundamentally because

if you have  autonomous agents going around on the  internet creating documents browsing web  pages and watching videos then get ready  because in the coming years you’d see an  invasion of these agents all over the  Internet and if you thought you had bots  on the internet before well that was  just the tip of the iceberg because now  we’re going to have real real Bots  working all over the internet now we  move on to the next big thing coming in  2024 and what we have guys is AI Warfare  so essentially the US Air Force is  moving pretty fast on AI piloted fighter  jets and according to the experts the AI  Pilots inside fighter jets would be  extremely effective for any Air Force in  the world this is an excerpt from the  article and it says here that once you  get through the process of connecting an  AI to a suponic fighter the resulting  maneuvering is endlessly fascinating we  have seen things that make sense and  completely surprising things that make  no sense at all thanks to our safety  systems programmers are changing their  models overnight and we’re engaging them  in the next morning this is unheard of  in flight control system development  much less experimentation with  unpredictable AI.

agents now take a look  at this video this fighter jet is being  flown by an AI pilot it has no human  pilot on board and despite that this AI  controlled fighter jet can perform all  sorts of Maneuvers it can destroy enemy  targets precisely and it can also engage  in dog fights with enemy planes recently  the most insane news came out that US  Air Force is going to induct this AI  system into six more f-16s now there is  a huge downside to having AI pilots  flying fighter jets what if these AI  systems get hacked or these AI Pilots go  Rogue because in that case it might end  up killing lots of people on the ground  and that of course would be awful but we  do have the fact that every time a human  life is on the line that is something  that we don’t want thousands of fighter  pilots have lost lives in the  battlefields so if you have ai flying  the planes it might help save a lot of  precious lives often times Pilots lose  Consciousness while maneuvering the air  aircraft it happens due to increased  gForce so the blood drains from your  head when increased amount of gForce is  applied to your body and you eventually  pass out and of course this is not good  because because you know if you’re in a  dog fight you’re flying around the  g-forces in order to maneuver if you  want to turn left or right really  quickly the g-forces can hit you and if  you’re not trained enough you can pass  out you can crash the Jet and die but  imagine having an AI system that doesn’t  experience g-lock because it’s just an  AI system it’s empty it’s not going to  have fear it’s not going to have any  kind of hesitation it’s going to know  exactly what to do this AI  pilot system  is being trained in all kinds of  different simulated battlefields with  different scenarios.

and once it fights  itself in a bunch of different battles  and it figures out what strategy the  enemy is going to use or how to dog  fight in the best way it is going to be  able to defeat all sorts of enemies in  all kinds of terrain it trains itself  the same way as alphago if you don’t  know what alphao was alphago was  essentially a board game in which the AI  played itself millions and millions of  times and because it played itself  millions and millions of times it knew  exactly what to do every single time and  developed unique crazy strategies that  nobody would have thought do you not  think an AI system if they managed to  keep training it just on flying jets  perfectly it’s going to be just the best  fighter pilot on the planet I think it’s  going to be absolutely crazy there’s no  way that humans are going to be able to  beat that kind of intelligence Bloomberg  posted this video of what they titled as  autonomous reusable fighter jet  this autonomous fighter jet can  literally blow up enemy drones so  essentially you can just deploy it from  a very small platform without any Runway  it would fly into any kind of war zone  and since it’s an autonomous jet it  would identify enemy drones all by  itself and start hunting them one by one  now think of it like this US Air Force  could have a network of these Jets  essentially like a network of them so  you can have like a swarm of these jets  flying around protecting Borders or  going into war zones and destroying  enemy targets with high Precision it  would also drastically reduce fatality  rates in Armed Forces as you could have  an army of these autonomous Jets  protecting you from the enemies another  thing that’s going to happen in 2024.

Leave a Reply

Your email address will not be published. Required fields are marked *