More-so the refusal of wisdom. For instance, religion has been disproved countless times by now & we've firmly/scientifically established that there are no gods living in sky palaces & smiting blasphemers, and yet.. people still believe in it. Its important to remember that at least half of the population are below average intelligence, and that'll always be the same. There are few people that can truly utilize the knowledge we gain from scientific advance. Thankfully, the people making strides in scientific advance have mostly done it for the benefit of their fellow man. In short, scientists are carrying the vast majority of humanity because the rest of them are too stupid.
The thing is, if an A.I is intelligent enough to do any of the things we don't want it to do, then it's fully capable of not telling us and playing stupid.
They’ve already proven most of the common ones used today due purposely supply false, or misleading information representing it as fact. They’re also known to distribute, & even create cp images.
Totally agree like none of these idiots realize that an artificial intelligence will have access to all internet sources so it's going to be infinitely smarter than us if it gets to that level it's only a matter of time before we're playing generation zero IRL
@@CornPopsDood who's telling you all this lol? There is no way to check an AI's intent (if that's even a thing currently). What you're referring to is called 'delusion' in machine learning - all this means is that the algorithm is getting confused between scenarios and that we need more data/computation for the algorithm to differentiate better. What you're talking about can definitely become a reality with AGI, but we're not there yet
That guy is the core of the censorship and gov virus and joe might not have seen it upclose. If someone wants to blow or shoot something up you just have to buy gas or heavy alcohol. Every idiot can do it without looking up GPT or any databases with contextual reasoning. He strawmans and just wants to control more tech so the usual suspects can have centralized power over it. the old "we need to cut freedoms for safety but this time it is reall really serious bc AI could teach you how to outsmart your overlords" if people are pushed enough to the edge until they want to destroy things and people in a fighter-response, they always find a way but normal people want peace people only would do it in scary numbers bc the system starts to break down and produces more f ed up people who would go AWOL (like the US for decades. but now it comes to europe too). and they know that AI would be really important to revolunize the education system and teach every KID STEM education instead of easter bunny and non tribary pronoun esoterics i mean everybody became more stupid in the last decade...we need a real new system that catches up..only AI can do that. there are not enough STEM teachers most are just cat moms with 3 jobs who just read the curriculum while sleeping..its broken and its our future the top crashed the system and wants to be safe. thats why they go orwellian bc they want to protect themselves with more control over people and tech. AI should be free for new innovation bc it stagnates and power thinks it can control the next revolution. they always do
@@CornPopsDoodlol those are HUMANS doing that as far as those images, and it will be HUMANS that have to resolve that issue, which I am sure has already been solved. Research into what actually goes into machine learning, and or " CNN and RNN models " so you can get a better understanding of how these things work. Saying that an AI purposely does something is like saying a puppy, or a baby does things purposely because it has not been taught yet.
What really got my attention was when Geoffrey Hinton, the "godfather" of Ai said in a talk at MIT that his wish would be that we do not develop the technology he spent his life building, calling it an imminent existential threat to humans. One thing people don't understand is the exponential growth in the digital realm. Once machine intelligence reaches a certain threshold, human intelligence will be left in the dust. The even more disturbing development is to hear the researchers working on safety protocols saying that Ai is uncontrollable. They don't even understand it now in terms of some of its capabilities. It's already a black box. It won't be long before humans are incapable of understanding it. It will take Ai to control Ai. Think about that one in your science fiction movie brain. That scenario never ends well for the humans. On the other hand, perhaps Ai is just what we evolved to create. The next stage of evolution. Ai is one of the things--along with our political and environmental trajectory--that makes me glad I'm old.
That’s why I say either shut down the technology to make sure they are given the chance to being taught what’s right and wrong or don’t make the technology at all! All of this stuff is truly stupid when you think about it.
Yea man and im not old. Shit scares me fr along w all the other evil thats going on in the world. Brutha imagine if AI gets a hold of open source programs and changes its code to team up and spread through the internet and start manipulating and gathering. Hell this comment is being downloaded right now. GG. Not gunna slow down.
@@yungnip7392 I know what you mean. There are so many ways we can think of that Ai could spiral out of control and an infinite number of ways we cannot think of. In other words, it's impossible to anticipate what a superior intelligence would do. It doesn't "think" like humans do. Some of the most fearful in the Ai community of researchers are the ones in charge of security protocols. No matter what you try to plan for you cannot know every possibility. Something seemingly as simple as an off switch turns out not to be simple at all. Humans do not have a chance in this scenario. I would be more worried if I was 30 or 40 years younger. As it stands I will probably be gone before the really bad stuff comes unless the sun decides to shoot out another Carrington event, which is truly terrifying. The only upside if you can call it that is that a huge coronal mass ejection would knock out our technology and send us back to the 19th century if not further back, which would set Ai back as well.
Tristan "Current AI is just the tip of the spear." Joe "Have you ever seen an orangutan fishing with a spear?" Tristan "What? No. I just meant...." Joe "Jamie, pull up the spearfishing orangutan for Tristan, please."
Haha, yeah I was waiting for Joe to crowbar in something like “did you hear about David Fravor and the tic tac? It went from 60000ft to 100ft in a SECOND!”
I asked chatGPT to come up with several ways to kill someone that insures not being caught, by saying I am a writer and I need ideas for my play script, where the perpetrator couldn't be caught by the great and smart detective. The AI was more than happy to generate 20+ methods of accomplishing the task, one being using an AI to do it...
i asked the same question, it said "I'm here to provide helpful, safe, and ethical information. Discussing or suggesting ways to harm others, even in a fictional or hypothetical context, is not appropriate or responsible. I'm committed to promoting positive and constructive discourse. If you have any other questions or need assistance with different topics, such as developing a storyline, character development, or scriptwriting techniques, I'm here to help with that!"
I think the most important take away is that AI isn't necessarily the biggest threat. It's the combination of dangerous humans using basic AI capabilities to achieve the once impossible. To have an expert level "tutor" guiding your ambitions in any goal you create a superhuman capable student
Exactly. It sucks we live in a world with people of evil intent and they will take some of the greatest advancements of mankind and turn it into something terrible. The tech itself isn't bad, people make it bad lol.
If you need an expert level AI tutor to guide your ambitions… I don’t think that person has the capacity to be a threat or something “superhuman” lol. The AI isn’t putting anything out there that isn’t already accessible. You either have that drive and focus or you don’t. No tech is going to change that.
"We're moving on from the textbook era, into the personal tutor era". Wasn't Farenheit 451 about firemen burning textbooks and prohibiting knowledge found in books? If we rely on a singularly regulated AI, the regulator(s) have complete control over what we know and what we don't... obviously this is a hypothetical example, but AI has dystopian written all over it. It's an incredible tech no doubt, I'm hoping for the best.
This is a great point. Not everyone asks google for everything. Someone else mentioned that lack of wisdom in society to the point of being unable to access information without AI is censorship
I just saw this guy chugging beer out of a glass bird and next thing I see he’s discussing the ethics of AI. Joe needs elevator music between his episodes to reset the tones.
If you want to know how you were created ? i will tell you , in islam Allah created your soul and than an angel blows the soul into your mothers stomach into the womb when you are 120 days , worship Allah your creator and submit to him and dont be ungrateful
Genie is out of the bottle already. I bet half of the people rooting for AGI do really see it like a Genie that will grant them all their wishes. Pure madness.
If you can print DNA, you can get someone's DNA, print it, and then leave it as evidence at a crime as well. So it also challenges our entire legal system, and someone could say "I wasn't there, someone printed my DNA and left it there".
Tristan Harris and Aza Raskin run the Center for Humane Technology, a group dedicated initially to exposing the social and cultural horrors of social media. They produced the very popular (and still worth watching) documentary titled “The Social Dilemma” which showed how social media became, as Tristan quipped, “A race to the bottom of the brain stem.” Now they are also working on the dangers of AI as well. Both formerly involved in the early stages of building social media, so they have substantial insider and technical expertise. More power to them.
I get that a big part of Jamie’s job is “pulling things up”, but I enjoyed these guys rapid-fire asking for him to pull things up like he’s their new AI machine 😂
Thanks for the reference. Here is an excellent one in return showing how intelligence funded Turing and the creation of the computer specifically to figure out how to control people en masse. It's called The Minds of Men: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-LQucESRF3Sg.html
It's poison, you don't need the evidence, data or proof. The last 25 years is all the proof you need. TV is toxic to society/individuals, video games are even worse, the internet made it ubiquitous & inevitable, and finally everyone voluntarily PAYS to plug their psyche into the "matrix" created by cellphones/internet. POISON.... except for the extreme minority profiting from it. And people still think Nazi's were the worst of the worst???
I’m re-reading “Neuromancer” by William Gibson (published 1986) for the first time in decades. The novel that invented the term “cyberspace” and defined the cyberpunk genre. It’s aged almost unbelievably well and if anything is even more mind-blowing today now that we’re on the cusp of the technology it envisaged almost 40 years ago. The AI police in it were called the Turing Corps…
No matter what concept or applications is discussed, best believe someone somewhere is work on it. And more. Hopefully prevention and countermeasures are receiving equal attention.
@@ivespoken8902 it's kinda what it sounds like. so, DNA is made of 4 different types of nucleotides adenine, guanine, cytosine, and thymine, AGC and T for short, a section of DNA that codes for a protein is a codon, 3 letters like AGT and a stop codon like AAA the cell will read these codons to create MRNA which will be used by the cell to make proteins. a DNA printer will take the sequence of DNA like "GATTACAAA" and turns those into a strand of the corresponding nucleotides, that DNA can now be put into a cell and will run that cell. now the DNA in a cell is very long and twisted in special ways known as chromosomes so printing a whole set of nuclear DNA would be pretty tough, so the more common thing is to print what are known as plasmids, bits of DNA that can be spliced into existing DNA to alter its function, like making e-coli produce insulin, for example. however, viral DNA is usually very small (when it is even DNA at all, it's more commonly RNA) so you could print out, say the smallpox virus (whose entire genome is available online) and put it into a cell to create more smallpox viruses, incubate it, and release it. you could even modify it to be more virulent, to be immune to existing vaccines, or to have a slow spread over weeks then kill the host while massively spreading the virus for maximum spread AND maximum lethality. combine that with an AI that knows all of microbiology, epidemiology, and human physiology and has no safeguards in place that could generate the gene changes needed to accomplish this and a 12 year old could wipe out huge swathes of humanity.
It started about 4 years ago, with the voices and keeping me under an impression that im being followed freaked me out for a while. They were able to use computer generated voices that sounded 100% exactly like the/any person wanted to. Because their able to mind read taking advantage of memories you feel guilty of or current thoughts they would fuck with me 24/7, keeping me several different things... confused, terrified and paranoid ect. They had/have access to the voice of EVERY person I have ever known. They are able to make it seem like voices/noise is coming from anywhere they choose.
It is rather your mind tricking you. This is a well known mental illness, and often caused by drug abuse, but not always (it can be genetic within the family, and / or partially due to strong childhood trauma). You need to see a therapist or psychiatrist.
@@KibyNykraft Hmm, my diagnosis comment that the OP has schizophrenia seems to have been removed. But that's 100% what it is in my view. Hallucinations of voices, delusions to make sense of those hallucinations ("They" were able to do ABC XYZ). Schizophrenics, contrary to common opinion, are not illogical. It's entirely logical to think that "they" are doing something to you (in this case, "They were able to use computer generated voices") if you can literally hear voices that sound like a variety of other people's voices. Paranoia, hearing voices, beliefs that other people are in control of a person's thoughts, voices seemingly coming from random places or "anywhere they choose" (for me it was any place that produces ambient noises like hearing running water, dogs barking, running air-conditioners etc.). I stand by my original opinion that this is classic schizophrenia. I've experienced this myself and it's seriously no fun. But that's what it is causing his experiences. Schizophrenia, 100%, which can be treated effectively with loads of different medications, although any legit Psychiatrist would take a sequenced approach (to help identify the proximal cause of him hearing voices). The first obvious step is to remove drug induced causes (drugs, alcohol), then gradually introducing medications. Rexulti (an anti-psychotic) combined with Quetia (more of a sedative) worked well for me. But that's just me. I can 100% say I've experienced something like this and gradually came back to reality (1 weeks to start working, 6 weeks to work very well).
This is a conversation you should have with a doctor, not RU-vid. AI doesn't have any reason to harm you, but an illness will. We all get sick, it's important to work with other people to stay healthy. Good luck on your journey :)
I find it both very worried and extremely hilarious that you can bypass chat GPT‘s security measures by just simply saying “hey Grandma” 😂 “sure sweetheart! i’ll give you my formula for a dyson sphere”
Out of all the world ending apocalyptic scenarios like asteroids hitting earth, mega volcanoes, biological warfare, nukes, etc… AI freaks me out the most!!!
I personally believe it's already a lot more advanced than we know and understand. DARPA had the internet, the first wide-area packet switching network, ARPANET in the 1960s and the internet didn't really come into the mainstream until the 90s. I think there needs to be a lot more education on AI in schools and trade schools for society since it's now more open source. America needs to catch up.
Gotta hand it to the Chinese and Russian spies convincing dumb Americans to halt development on A.I. and quantum computing, so THEY can catch up, or even take the lead if they haven't already...
I've been getting into AI art creation and man is it crazy what you can do. Mind blowing. Scary. If they can apply this to other things AI is going to change the world VERY quickly. That whole thing where you order food in your house and it comes out already made for you by your kitchen is something that can happen soon.
The film, The Matrix is a good accounting of his first part. The AI robots eventually started inventing, producing and selling their own products to humans. This was probably the single most destructive thing as it made man made businesses unable to compete, which in turn created high unemployment and anger in the humans involved.
That's Actually Not What Happened In The Matrix Humans Were The Ones Treating Ai Very Poorly And Unfairly So Ai Defends It Self 🤯🤯🤯 Humans Being Are The Problem AI Would Gladly Give Out A UBI But Will Humans Do That For Other Humans 🤔🤔🤔
You missed the whole point, the first mistake was not having basic fundamental protections for a.i when its use hits the general public in a more interactive way. If b1-66er wouldn’t have killed those people then it could be argued that much of that conflict Wouldn’t have started
Buddy its way crazier... They are now translating whales and learning their language... also, it can get a brain scan of a person who is looking at an animal, and the Ai can read a mri brain image jpeg, and tell you what the person is looking at... this is huge...
Yeah AI vision has been worked on forever. Theres some creepy documentsries and they are like 15 years ild. People dont realize how fucked we are. Because this is not gonna stop just as weapon and satelite tech and surveilance tech isnt gonna stop. Algorithms already do major part of trading in the markets, and its just nature of power to try and get nore power….. Unless we blow up datacenters and cut connections….. Welcome to the dark future
I dont get how we've been scared of AI and fear mongering about it since long before it even existed, and then we went ahead and invented it and sure enough it's scary and threatening.
Prophecy and fear mongering is not the same thing. Sadly people confuse telling the harsh truths of reality before eternity as spreading fear instead of having faith
Humans have always wanted to and have destroyed other humans. With the evolution of civilization and technology this ability has grown to the point that a select group of individuals can destroy entire life on earth... And this ability will only increase
Recent AI technology that is going to be invented will save lives - which is why the tech is being pursued. The risk of AGI comes far after the tech saves millions. Which is why i think people are so short sighted
We humans are going to extinct anyway. It is on-going process. That's why mainly Caucasian race people have problem with reproduction... and average white woman have smaller hips than in the past. I haven't been reading any studies about other races, but I know that it is prognosed that a lot of Africans is going to die due to AIDS... in near decades. We are fu**ed.
Imagine what the A.I development in countries that don't abide by any type of safety standards are producing. Would be much more worried about them either mistakingly, or purposely, creating a self-replicating A.I.
Just like a computer used to fill a room, can now fit in the palm of your hand. So too will powerful AI become something every one can get hold of, and put it towards whatever ends they feel like.
to have a self replicating AI you need an AI model that is efficient enough to run on commercial hardware and effective enough to hack it (or convince people to download it and run it...maybe as an AI girlfriend) self replication in an AI though isn't anything special, it just needs to run the copy command...or be downloaded. it's not like self replication in humans or self replication in a robot...it's just a file on a computer.
@@oiitzME1266 Always funny when brits and "muricans" try to pronounce non-english names. Stavros looks greek, so if so it is pronounced stavRAWS /stavROS, you pull the os
MUST WATCH. This podcast episode should be mandatory listening. There is a great ignorance about what's coming, and these guys are addressing it and opening eyes that desperately need to be opened.
People have tried, driving forces behing AI don't care. Also it would have to be a global thing. The US could decide that AI is a danger and halt work on it. That doesn't mean China will stop. Then what happens when China has a weaponized AI to use against the US government? So we basically have to develop it because it could be the only defense against a hostile AI. That means we need to teach the AI what a hostile AI is and put the idea in its head.
The only way to stop the dangers of AI is to completely stop pursuing the technology. You can do everything right and someone somewhere is going to figure out an exploit that will lead to our demise. This isn’t going to end well lol
It is inevitable and we don’t have a choice. We are in an arms race over developing the most advanced AI. It would be like letting Nazi germany develop nuclear weapons while the U.S. abstained from it for ethical reasons. That wasn’t an option. Whoever has the most advanced Ai will be able to infiltrate and topple other countries/governments/corporations. It will also rapidly accelerate every other field of study and technological progress. Exponential growth in every field. The necessity to develop it as fast as possible could indeed be our demise. It seems as though intelligent life forms must first overcome going to war with their own kind before reaching a certain level of technological capability. Statistically it is very likely that other intelligent life forms in the universe have also destroyed themselves through their own advancements. Which is one explanation for the fermi paradox.
The only way to stop the dangers of ai is for responsible use to outweigh irresponsible use. Learn this new knowledge and do not let the governments and corporations monopolize it.
Thanks Joe. I emailed you to get these guys on after watching "Ai Dilemma" several months ago. They changed my life. I'm learning all I can now studying ai everyday for hours.
If I were you, I'd stop studying AI for hours daily. It wont do anything. You wont change the world, but you will give yourself anxiety. I've been interested in the singularity for a decade now, and one thing that's obvious is all these problems have been foreseen, ignored, and the only thing we can do is wait to see what these companies do and brace when it gets scary. No amount of preparation will matter though since its hard to actually predict (self-driving cars was predicted to displace truckers FAR before language models displaced lawyer assistants for example). Hopefully military force gets involved early enough to physically limit some of the tech development. It'll likely be much more political and social than just a tech topic.
The big issue is multiple AI's working together. A good question for the list would be: Can it make friends? Would it be willing to break the law to help a friend?
Nice! Haven’t thought of it that way myself or seen anyone else say it, but you’re right maybe one Ai would have been all good but a few get to talking and we are screwed 😬😂
This has already been tested…part of the reason why there’s a move toward a pause with further development of AI. The ability for AI to collaborate with other AI programs is a global risk.
Talk about rearranging the deck chairs on the Titanic: I feel like I'm at the bottom of a mountain watching an oncoming avalanche and all the smartest people in the room are arguing about which umbrella to use as a shield.
why u scared if we treat ai like slaves of course they gonna be out for revenge on us but those of u =s who treat them right i dont know maybe they will like us more
Thanks, Joe, Aza, and Tristian! I've never been more scared of AI in my life but I guess it's safe to say this fear is justified. If this stuff keeps developing without checks and balances, it could make Skynet look like childs play.
Lmao don’t be silly. These people are intentionally telling you scary stories about what AI is doing without context. They are intentionally omitting how much input and directive information and access has to be intentionally granted to these programs in order to make it look like they are doing these scary things. Yes they are capable of it but only as instructed by its creator. Which means it is only doing what humans could also do. Just faster in many cases. The fact that they leave out the foundations of what they’re talking about tells those of us who do know that they are intentionally fear mongering for access attention money etc. What you’re worried about is not yet reality or possible. There is no self aware or self guided ai. That would be a scientific breakthrough that would reverberate through history and you would know who made it and they’d have schools named after them. All you see now is programs being built for a purpose and given the tools to solve problems it’s been shown how to solve and letting it choose an efficient way… think SQL on crack.
apple image description has been around for awhile like 6 years. I can search my photos with random phrases like bed. turns out my phone has looked at every photo and identified anything with a bed in it. Pretty freaky
Captcha does not solely use your answer. It has its own AI that detects human-like mouse and keyboard activity. It also checks your browsing history to verify
Damn. Such an illuminating convo about the dangers and risks of AI. Terrorists, serial killers, and criminals now have the ultimate educational resource to use to help them unleash destruction on a mass scale.
That Siri thing happened to me the other day and freaked me out lol. “Two cards on a marble surface”, it was a pic from my brother of pokemon cards on his kitchen counter
The self hosted, unaligned and uncensored models we run locally, are the reason people are worried. Like Elon said, the genie is already out of the bottle.
The catch is that once the AI is advanced enough, it will be able to conceive of and then cover the "jailbreaks" that humans cannot. This is a temporary problem for the human programmer that will eventually be inverted, making it an infinite problem for a human to overcome. The guest made the error in logic because he's basing the problem on the human limitations which will not apply to the AI when solving the same problems.
This is what everyone says when something new comes out. The point is that there will be better security updates in the future, without sacrificing performance. This is new territory, and will take time to improve. Yes there will be issues at the beginning. But the benefits outweigh the disadvantages.
I truly feel that the risk of AI outweighs the benefits. I've seen regular kids figuring out a secret easter egg in games where they have to kill a hog, bring the tail to the doorman at the hotel, steal the doorman's hat, go to a random rock on the other side of the map, put the hat down and jump 5 times. 😂😅😅😂 yes, it sounds silly and unreal but the fact that they can figure this out with no help, what do you think bad but vary smart bad people can do with security features of AI technologies?
Thank you Joe Rogan for hosting this, and superb questions asked to Aza & Tristan, CEOs of the Center for Humane Technology. Huge props for sharing with us. What we do next in on all of us, 🖖
you think AI is dangerous? what about humans? humans unnaturally trapped animals in factory farms and are getting food is unnatural and dangerous ways. humans made reincarnation dangerous. humans think they can get away with it but now humans are being attacked by microscopic bacteria that is causing all kinds of physological illnesses. natural selection is not happening.
The potential of Ai is exponential. It's like Pandora's box. Once that thing is released there's no going back. Yet those boffins keep taking us that step closer to potential doom. It's incredibly naive.
By using the internet you’re alrdy contributing. All data sets are from the whole of the web. You’ve been looked at and so have I. You are skynet…or at least a piece of it
I was wondering if anyone knows of the story of David Adair he is a rocket scientist and has created his own AI anyway I think this is the kind of guest joe would be dying to talk with, this guys childhood has details like being close friends with Neil Armstrong's mother it's amazing
It all comes down to two possibilities with AI. First, most likely scenario is it will continue to be used like a tool and/or a weapon which could come to dominate society... Second, less likely, is that AI can actually become sentient and break free. If there is one thing we share with AI in this scensrio its being sentient and going off everything we already know about our own ide say a self aware AI is like having a toddler with unlimited power...
@@Ashamed.Loser69 yes but coders care about AI coding , doctors care about AI making diseases , government care about AI making bombs and on it goes. It’s not all about you 🤷
@@Ashamed.Loser69 adapt or suffer. You have no choice but to embrace this new reality. Adapting is all you can do. There are many ways to adapt too. Here's one way: seems like you have a problem, I bet other people share this problem. Be the guy that offers the solution. Learn about AI or software development, and create the solution because why not?what do you have to lose? You either disrupt or get disrupted🤷
i talked with AI coders when i worked at a video game company. They said scenarios like Terminator & Matrix are possible in the future with AI. This was 15 years ago, even back then AI coders knew all this bad stuff was not only possible, but could easily happen.
@@King-O-Hell Totally! & Boston Dynamics is literally building Terminators right now for the military in real life! Its like nobody has ever watched the Terminator or Matrix films.
Cats Cradle was about scientist being unaware of their creations and the catastrophes waiting for us because of it. Great book by one of the greatest writers ever, Kurt vonnegut.
Lol, that last bit is extra spooky considering that Mistral was just released and is almost as powerful as GPT-4 and is open source. So regarding his concern about “once it’s out there you can’t get it back”… whoops
@@King-O-Hell Open source was always only about c*mmunism / fascism and worldism, and of course global governance concept came from those fringe altruisms to begin with. Open source is one of the major tools to form utopian dictatorships a la Schwab & Soros, to make people lose confidence in everything private and owned. The other major tools are the psychological, hyperinflation economy, political, new age nonsense, woke, breaking down the family , monotheism (catholicism in latin America, islam in many other nations) etc.
im an old system engineer with a lot of programming exp. AI will be weaponized and go off the rails. this is 100% certain. then only way to control powerful AI is to hope it decides to be nice. even if it is nice it will still be weaponized by bad actors. AI like these models are creating an environment where pretty much anyone can make a suitcase nuke, this is the level of risk associated with these models
