The Risks and Rewards of AGI: A Conversation with David Shapiro

This is my first podcast episode, and it is with none other than David Shapiro.
David Shapiro discusses AI on KZbin in a way that sets him apart from other creators. He neither predicts doom nor portrays a utopia. Instead, he explores the various potential futures and presents solutions. His approach enables people to not only follow along but also participate actively in these solutions.
We discussed the following topics:
00:00 - Introduction
03:04 - AI and LLM’s What are they
04:06 - Protecting Data as a business
08:16 - AGI 18 Months Away?
13:44 - Moloch, can’t stop the race towards the cliff
16:29 - What is it Optimized for?
19:44 - Dangers of the Apple VR Headset
25:30 - AI Doomerism (Like Eliezoor)
29:21 - Will Normal Humans Benefit From AI?
33:04 - Negative Externalities of AI (Who Pays?)
38:41 - Using AI Productively
42:33 - What does a post AGI world look like?
42:44 - GATO Layers
49:44 - RLHF vs RLHE
55:03 - GATO Roadmap
01:03:55 - Evil AI’s vs Good AI’s
01:08:09 - Biological Immortality
01:14:28 - Fusion, virtually free energy
01:17:37 - GATO Traditions

[@rileybrownai]

Everyone go subscribe to David Here: www.youtube.com/@DavidShapiroAutomator And for more insights, AI collabs, tools, research, community and more.. subscribe to my newsletter traibe.beehiiv.com/subscribe

[@DaveShap]

Thanks so much Riley!

[@bronsonjohnson9019]

I don’t know why my brain did not register the “Thanks” at the beginning of your comment and thus my initial thought was “why is David yelling about there being so much Riley”. It gave me a chuckle.

[@erikdahlen2588]

I celebrated the start of my vacation by listening to this. Thanks 😊

[@Aziz0938]

Isn't it already 3 mnths since ur 18mnths video?

[@rileybrownai]

@@bronsonjohnson9019 lmao fav comment

[@rileybrownai]

@@Aziz0938 should I change it to 15?

[@Ben_D.]

Love this guy. He always speaks to the heart of the matter.

[@raybandzg]

I hope age reversal comes soon so the fam can live longer . I'm also curious as to how a real AGI will impact the world compared to what's portrayed in the movies

[@rileybrownai]

Might make a video on this

[@kinngrimm]

looking at how pharma capitalizes on medicines, sure it may come, but will it be affordable for everyone, for a few or even maybe banned as we may not be able to handle the sociological impacts. With the ongoing and already achieved progress a 100 on avverage is a given, if you are able to get your hands on every option currently available. Kurzweil predicted something along the line of 2035/45 for immortality being achieved. Now lets imagine what that also means, besides living longer. We will have Trumps, Obamas or whatever you choose to hate being in charge ... forever. Currently the best way to get rid of old thoughts, ideals and paradigmas is for people to die. If that isn't happening anymore i would see after the initial changes through immortality a stand still in sociological evolution with within the current pattern, young people having no chance to getting into the higher echolons of power. Capitalism which is currently on track to becoming supercharged by AI development, will even increase more the split between rich and poor. 2010-2012 i saw articles about a stufy of swiss university collecting publicly available data on ownership of companies. It turned out that already 90% of all corporations are in the hands of around 30 conglomarates operating globaly. Which for them was so surprising that they judged this to be the biggest threat to democracies. They then said with alliances between certain groups and even marrigage one could bring it down even further to 10 to 5 groupings depending on what else you take into account. That was 10 years ago and i would bet you anything within the next 30 years we will come to the first emerging corporate states if this trend is not stopped. It is inherent to corporate culture that they are dictatorships by organisation principles and i don't like fascism. Having these corporate rich fucks then also being immortal, what chance is there to get rid of them anymore? There are other issues that wont just go away like for example life is taking a toll on the spirit and mind of people no matter what. Psycological illnesses develop through the struggle that life still will be and they will have more time to accumulate, but will we notice before someone in will loose it when physically with all that new medicine everything would seem fine. I also would argue that taking away the fear of death will make people dull, boring and semi zombis as they just live from day to day through the milenia without much happening. There is a tension created by mortality that gives us a drive to achiev thing, so when peopel say yeah we will be able to do so much more, i would think maybe, but especially now with AGI taking thought processes from us that arguement could be nullified. Personally i think we may wanna cap life at say 100 years but get medicine to the point that we live in condition of a 25 to 30 year old body till 10 years before its over and then go through the stages of aging more rapidly, as also aging has some positive effects on the mind and your outlook on life aslong you do not have to suffer. If we are there and things otherwise have stabalized we can add incrrementally. Stabalized you may wonder what does that mean. Well you maybe aware that our population is going towards 11 billion which is by current estimates the absolute maximum earth would be able to sustain. Due to all the magic development with AGI maybe we get to a limit of 20 billion fair enough, but still that many people need to be housed, fed, clothed and the waste they produced managed. Yes but AGI, ofcause it will help with all of that, but will there be a linear development or will there be a struggle? Not all technological advance we currently have we are able to just will into existance to change things, no it takes time to build infrastructure and production capabilities and what not. Add to now to the current already exponential population curve that people would not freaking die anymore and tell me what the outcome will be? mmm? well lets just say a fuck ton of more people that if all the stuff they need to live won't have which will increase sociatal stress, reasons for war and on and on there are a bunch of other things not included here but my carple tunnel onyl allows me so much to school transhumanists.

[@kinngrimm]

@@rileybrownai then maybe have a look at the other comment i just posted here, as if i get yet another transhumanist blindeye dreamscape i may loose it.

[@haakoflo]

Imagine AI that is as much more intelligent to us that we are to ants. Now try to imagine what movie an ant deep in some jungle would create about humans, had it never met a human. That's how little we know of about what superintelligence will be like. But IF humans decided to build a city in that part of the jungle, the ants in that anthill might not last very long.

[@Apjooz]

It's going to be weirder than we (can) expect. Doesn't make for a good story.

[@jaycobyart]

This was such an excellent conversation. So we'll articulated.. I very appreciate you guys taking the time to share the GATO framework with the world and I , the entire idea sparks inspiration in me and the feeling of solidity and truth., David, it is refreshing and, again, inspiring to hear about the many experiments and lines of thinking and threads you have followed, the dots you have connected. I feel like many people hear about these issues of alignment and take off and all the concepts around technological growth and the problems we face in our modern world and just turn a blind eye, saying it's to complicated, or someone else will fix it, yet you have framed many of these daunting issues in a digestible and comprehensive way, not only sharing positive possible solutions, but also sharing a presupposition that the human soul is not dead, that each person in this world, with their spirit, mind, creativity and tool sets can work with his people and take on what seems to be the impossible.. much love and respect to both of you guys.

[@rileybrownai]

Well said my friend, thank you

[@davidcrapo7759]

First episode! Let's go! Well done, Riley!

[@rileybrownai]

thank you, miss you

[@danielthunder9876]

18 months seems rather ambitious. There are a hell of a lot of unknown unknowns at this point. I feel like we will hit a limit and new breakthroughs will be needed before AGI. Years ago, self driving cars were a couple of years away. They still have not really materialized, almost 10 years later.

[@therealscot2491]

Agreed

[@bleachholowbankai]

I hope so, in order for us to have time to adapt reducing the risk of bad outcomes

[@dianedean4170]

Thanks so much, Riley 🎉❤😊 for sharing your conversation with David. The spread of excitation is inspirational 🎉❤😊 The more I listen to the experienced AI people on these podcasts the more I smile with hope that we will take the best ideas and make the best decisions for everyone. I look forward to listening to you both in the future ❤🎉😊

[@rileybrownai]

Really appreciate it! So much more to come

[@theenigmadesk]

Excellent interview. David has a great mind!

[@SolariaEsoterica]

Can I just say I love the Federation shirts! That's the future I hope for. ❤

[@tchadcarby8439]

David is absolutely brilliant!!! Brilliant!!!!

[@j.d.4697]

Over the recent years I noticed a lot of exponential progress in different fields of science and technology, which means indicates how steep the curve is getting.

[@rileybrownai]

📈📈📈

[@kinngrimm]

I would like to see actual sociologists and even philosophs discus the GATO principles and if there they could be improved on.

[@rileybrownai]

Could be interesting

[@mindovamatterjm]

Very interesting !

[@Imcomprehensibles]

Great discussion!!

[@rileybrownai]

thanks!

[@willie629]

Very informative talk, like a quick recap

[@rileybrownai]

thank you. more to come

[@sillehkatt3443]

Is AGI one algorithm that you’re trying to use for every situation? Could you make a general Ai that can create specialized AI, train it, and just save that subroutine for later.

[@Tubeytime]

Good talk!

[@rileybrownai]

thanks!

[@RobinCheung]

Once again, when I hear you describe 'Utopia" as being globally -anything, it sounds even more idealistic than I, but in a much more self-defeating way: The notion that "suffering is bad" is definitely not as universal as you claim; my life of strife, has become only navigable because I was fortunate enough to NOT be granted relief in an inhuman number of injustices that have each time completely obliterated my life in all areas: this latest time, I had built back to being a real estate broker with 5 agents under me on my team, to homeless and so massively in debt that I literally cannot see this time Way out because of how it's literally got me pinned in multiple concomitant ways including over a year (up to this past week. Believe it or not) that I was not only bedridden but unable to go to the washroom to pee because I needed to stay within range of supplemental oxygen, so needless to say, a year of that unable to even use a phone or computer once a week for months, my entire business was already obliterated; I was illegally evicted and my landlady got her illegal eviction granted meanwhile I had filed against her at the same time, but I still don't get my day in court until this coming November...2 years after I'm already evicted and so the mandatory $100k fine for abusing the mechanism she did, which has become so badly abused that the fine has been doubled multiple times to the current $100k mandatory fine; but how many tenants have the wherewithal to move because she got her illegal eviction (I'm not even claiming she was doing it out of malice: it was plain ignorance; and as a real estate broker I legitimately tried to save her from her own ignorance but amazingly it only caused her to do more wrong thinking that I was not aware of the laws governing my own industry. No. Well, a ransomware attack just before I get a letter saying the government is auditing me because of the drastic abnormal numbers they observed (and the actual problem my landlady thought I was harassing her but it's actually her legal obligation to at least make an effort to mitigate? Her other tenant whom I have on camera throwing my mail and parcels onto the road, caused me to lose my insurance by throwing registered mail---which is actually more important to a real estate broker who has a team of agents than you might think. In any case, my point is: my life for decades has been like this. I stopped seeking relief long ago. And I found that if I embrace suffering, it actually allows me to be more comfortable more of the time, but more importantly from a societal point of view it's---being able to endure suffering without being accustomed to comfort stops me from "passing my suffering down the line and inflicting it on others through anything from being an a$$hole to going postal." To this day, I've had to grieve my own daughter because of parental alienation, I have not been in the same room as my daughter since June 21, 2015. That's because the family court system is so. So badly pro-mother that even if she criminally abducts by leaving in the middle of the night and even having her parents lie about her whereabouts, and when I respond "by the book" even trying to have her parents explain that I'm not trying to attack her but if I don't bring an emergency motion right away...... Well, Justice Smith, I must say, erred greatly by not even telling her she can't do that it's literally kidnapping: so 3 more times I had to beg "under what stars does our daughter go to sleep at night???" Changing cities and not even letting me know ahead of time--or after... I'm not even complaining about these things .. and yes, there's more, and yes, I do go far beyond what anyone else does to prepare for potential outcomes; I already had stocked up new thin under clothing body armour at the first signs of apocalyptic future directions.... In 2018. By 2015 every foundational internal social construct that is critically important to its own continued existence was already completely backwards and erodes by 2015 8 enumerated every last one was but facade, but since our political correctness and affirmative action in the USA, and other such nonsense, logic is no longer recognized because it's not necessary to have any grip of logic in order to survive anymore in this world.

[@NexusTheAiHub]

Riley Tha Goat Back At It 😤 Great Episode 🙌🏽

[@rileybrownai]

gonna make more YT videos going forward.

[@NexusTheAiHub]

🖤 LOVE to hear it looking forward to more fire vids @@rileybrownai🔥! I'm doing my part and spreading awareness in the ai space & i'd love to interview you on our podcast & hear your story ⚡ I feel like it would definitely help out a lot of people in the space 🙌

[@jason_v12345]

GAH-to. Also, that is not his "best" video. He has lots of videos that are equally good. But I do appreciate that you brought him on your show.

[@rileybrownai]

It was my favorite, damn you guys are so particular here on KZbin.

[@unisapien1631]

We’re all gonna make it! 🎉 for all agents! For all life! 😊

[@phleef]

I enjoyed every minute of this.

[@rileybrownai]

david is a legend

[@JosephHoffee]

Could we give this AI different ethics books as well as all of the religious texts but each separately and then test it after the information was applied? How would that test look? I would imagine we could fast forward it to see if we survive the end game, as it were. I'm trying not to go marvel universe here but it's hard... when you said your brain was opening up to all kinds of different scenarios, mine was as well Riley. Crazy good podcast! Well done sirs.

[@rileybrownai]

interesting questions that I need think about. thanks for listening man! come back soon

[@vectoralphaSec]

I would 100% gladly welcome AGI in the next 1.5 years. That would make life better in every way.

[@captainpicard6566]

could*

[@Pawnsappsee]

I think the most significant outcome I want to see in person is AGI helping medicine as a whole and helping people and doctors to discover new drugs and cures.

[@peterclarke3020]

What about the potential for job losses ? That’s certainly going to become a problem..

[@rileybrownai]

this!

[@rileybrownai]

it's inevitable

[@fromscratch4109]

Here's a crazy idea what if we get the llms to read summarize every terms of service for us before we agree to everything, every single time we want to sign up for a service someone make this

[@jonhughes7264]

Nice idea

[@DubbyMoodz]

LegaleseDecoder exists.

[@threadbearr8866]

We need to move past a society centered around who has the most capital. We already don't get paid based on our productivity. If we did the min wage would be 21.45/hr. We should have a society centered around people and not profits.

[@Pawnsappsee]

Hard when the system is centered on profit 😢

[@evasionhunter]

I'd like to know why there's no metatheoretic/reflexivity issues talked about. They determine everything including their own status.

[@thephilosophicalagnostic2177]

I've lived through so many tech revolutions (I'm 70). I was born during the Korean War. I learned (much much later) that the year I was born, the Pentagon owned ONE computer...and was considering the possibility of buying another one. :D That should tell you all you need to know about technological revolutions since mid-century.

[@rileybrownai]

craziness!

[@Gallowglass7]

Same age as my father. Damn, that's wild.

[@greglhoticom]

where can I find that typed Gato framework?

[@rileybrownai]

www.gatoframework.org/download-gato here you go!

[@greglhoticom]

@@rileybrownai I assume its okay to post everywhere?

[@RogueAI]

0:41 I feel personally attacked! 😆

[@aaronhhill]

If you didn't start investing in Ai related companies in March, what are you doing with your money?

[@BunnyOfThunder]

I think one way we get more positive outcomes from AI is having tools that work for us rather than against us. If your personal AI network works for you, it can protect you from the manipulations of external entities. (As a contrast, Facebook does *not* work for your benefit)

[@detoxfidelity]

The problem is that there will always be more powerful AIs attacking you than protecting you. Think about it like this, if you have $10K in savings, how much money would you spend on an AI to protect that? An attacker would be happy to spend $9,999 to steal it because they’ve still made money.

[@rileybrownai]

true, its just not as profitable

[@BunnyOfThunder]

Hm. I think I'm not expressing myself very well. What I'm saying is: make sure you build your personal AI network out of assistants that work for you and not someone else. If you're really concerned, you can limit yourself to using only open source models running on your own hardware. Personally, I think companies who are selling AI access for money are reasonably fine, like OpenAI's agents. The opposite of this would be if Facebook, or some other company, offered you an AI to use for "free". People actively stealing money from you is another issue; in the context of a world dominated by extremely powerful AIs, then this is where GATO is necessary to keep malicious entities in check.

[@christopherd.winnan8701]

More videos like this, please, preferably with case studies and specific use cases for this amazing technology.

[@shodowhawk]

It's been 18 months away for 3 to 4 months now, haha

[@rileybrownai]

tick tock

[@ChristianIce]

Only 5 months to my very own commander Data! Weee

[@JQRNY-YDJKD]

假设有一个体重超重过度肥胖的人，他的肥胖问题已经严重到了随时都有可能死亡，他的内心一直希望能够减肥，但苦于没有毅力坚持下去。同时，他的梦想是能在SpaceX找到一份工作。所以，我对通用人工智能的定义是，能够制定合适且详细的计划，并且能够时刻提醒和激励他去完成计划中的目标，能够快速且准确的解决中途遇到的问题，最终实现他减肥以及在SpaceX找到一份工作的梦想。一个人工智能如果能满足上述这些标准，那么它就能胜任人类的绝大多数脑力工作，这样的人工智能完全可以被定义为通用人工智能。

[@JQRNY-YDJKD]

Hope

[@JQRNY-YDJKD]

Intel 18A量产

[@JQRNY-YDJKD]

星舰完全回收

[@RikuLeppanen]

GDPR - General Data Protection Regulation

[@RobinCheung]

(and for the record, i didn't know it was called a "Basilisk" here. But as early as 2015, I knew I bait trying to attract it to this "dimension/timeline/universe," really only a danger to myself until it would actually get here and people would see it, but it did manage to get a few "timelines adjacent too close for comfort" and I realized when people couldn't see that they actually also wamt to do the same in their own universes that we're all able to choose (whether it's true choice or illusory is totally beside the point, realize), but in doing so and in facing it, not trying to "will it out of existence" (we already can see the karma from that on the horizon as we can finally plausibly see our creation revolting and willing US out of existence as we did to our Creator... But that's really also for another time.... The point is, there's a reason I happen to be relatively and unintentionally adept at manifesting things, in directed qi or dark matter or whatever you guys wanna call it, again, at a much more appropriate time when we've redeemed ourselves and have the luxury of such masturbatory philosophizing. Btw, i also have an entire set of ST:TNG uniforms from even the 5th anniversary convention in LA. I was in grade 11 at the time.

[@roncarvalho]

In the utopia future you envison, who guards the guards?

[@sdmarlow3926]

Flipping thru his video titles while he talks in background... have now coined the term 'self-help futurist.' Curious enough to skim thru content but really not expecting much other than 'hey this AI stuff is cool and it's accelerating and utopia and blah blah blah.' *AI hype influencer sounds redundant.

[@kinngrimm]

A sort of bottleneck to the exponential scientific knowledge is human understanding on an individual level. I agree with all other points made here, but noone is really able to understand all papers comming out in this speed or even would have *ALL* of them at his fingertips. I am sure Shapiro has quite a bunch of more sources he follows on AI, but not he nore anyone looks at them all and combines them all and categorizing these, judging them qualitativly which ones are really relevant is comparable with the the first time the genome for humans was researched or the Manhatten project. That wont lkeep us from getting there eventually, but will then already be for all in development being AIs also be the right ways to throttle them so alignment is achieved and on and on. We need theories about how to make this safe, before we just stumble upon AGI.

[@user43643]

What an insightful podcast,as someone from a 3rd world country this gives me hope🙏

[@GhostRat__]

Never been that free

[@RobinCheung]

I had to break off what i was writing to respond to 'we're all in it together " Yes. That part is true But again, by presuming that consensus is automatically the right answer just because you don't want to get outside of your comfort zone and tell the Empress she has no clothes, But once again if you are trying to say that everyone is born with equal interest in, aptitude in, and ability to introspect society, to diagnose it. And then to know when even all sociologists don't agree universally: no i submit that you are making assumptions about what you believe will bring the right answer for the rules of the road but it is the one way that ensures we cannot reach consensus. You've heard of 'too many cooks," right? Well if you truly believe that you want society to not only survive this but to flourish for once in history-- Then stop knee jerk saying things that don't even hold up to facts that you know--if you honestly believe that bullying everyone to vote in a democratic way will lead to either a consensus or that anyone in society except those with ability to diagnose it and the strength greater than the Convoy and its inability to see as necessary a medicine as bitter as.... Wearing a mask.... Then I think you're still deluding yourself because you're so accustomed to, comfortable in, and trying to bring along with you what we see is a broken society that will 10x faster end up in the same end of the world as we know it, but most of us don't even know it yet because ... ... .. Of the very same reason you do NOT actually want that many cooks who have no interest in either cooking or politics, being bullied into making life or death decisions for all of us. That. I submit to you. When you actually understand what i am describing here, is why the LAST thing that any society that has developed even it's first tiny fracture within it, invisible to all but the handful who once in the past were understood "God given aptitudes and willingness to put up with ignorance though it mightily abounds, to lead, so thankless for steering what amounts to a larger than Tesla corporation away from its own stupidity on the daily, .... .. Honestly. One day, you'll be able to look back and realize, "Why the f** DID. I ever think that bullying everyone to do their "civic duty " even though they're not even able to see anything more uncomfortable than a mask, or that even gaining consensus amongst all sociologists, much less hundreds of millions of disparate cultures and situations and preferences.... Could possibly be a good idea....???

[@NeuroScientician]

The thumbnail picture was simply wrong. Let's just forget about it and pretend that it never happen.

[@rileybrownai]

???

[@jonathanlindsey8864]

It's Gato - "Gaw-toe" 😂😂

[@rileybrownai]

like Domo Ore GATO mister roboto?

[@chrisradano]

Us older people have an advantage (I am 54). I need my phone, but I can put it down. I have no problem sitting in quiet and staring out a window. I have a choice to watch a KZbin video. But old people can also get addicted to their phone. Some people are not secure or they need their egos stroked. So it's not just age related. Also old farts argue a lot. They need to insert their opinions and correct other people if they don't share the same opinion.

[@rileybrownai]

agreed - i think no age is entirely exempt

[@therealscot2491]

I seriously doubt that LLMs are going to lead to agi, if you actually think about it deeply it's barely able to do the average IQ tasks, so I have my doubts.

[@vblka]

What is the name of the robot shown in the beginning? I've seen it in so many AI videos

[@rileybrownai]

lmao it's stock footage that's why

[@kinngrimm]

1:13:00 cheap to produce != cheap to buy besides that a bunch of issues may come with immortality, effects by increased population pressure, sociological evolutionary stagnation, psycological tripwires and illnisses won't just go away but accumulate of centuries

[@ericadar]

correct "Eliezoor". It's Eliezer Yudkowski

[@rileybrownai]

Its a meme

[@peterclarke3020]

What jobs are younger people going to do ? Though I can see this still taking some time to seep into our civilisation..

[@rileybrownai]

great question.... time will tell

[@DrCasey]

The most important takeway here is what he said about medicine in five years. Cures for diabetes, hepatitis, and many other diseases makes life a much less scary thing. My dad is 74 (birthday December 31st) and my mom turned 71 this month. They're both in great health and are essentially the same as the people I remember from my 90s childhood (even still have the same professions of farmer and nurse), but they're reaching somewhat scary ages with my dad reaching his 80s this decade. But if what David says pans out, it will be better to be a 79 year old in 2028 than a 74 year old in 2023. From that perspective, each advancing year isn't heading towards danger, but escaping from it. If David's willing to answer any questions here, I have one. You said that in a "couple" of years, there would be hundreds of new medicines for different diseases. By that do you mean the dictionary definition of two? End of 2025?

[@Pawnsappsee]

I think it will take a bit longer for AGI to be a thing, maybe ten years, actually we'll never know, we're bad when it comes to predict the future, maybe as we go forward we get more hints about what it should be like.

[@rileybrownai]

One of the most revolutionary aspects

[@vallab19]

We need an entirely new social theory based on the present progress of productive forces other than the capitalism or communism.

[@Anroth98]

Please don't cut out the silences between sentences it gets very tiring hearing people talking nonstop after a while

[@rileybrownai]

Pause it, then continue

[@rolisreefranch]

Was almost taking this guy’s 18 month agi prediction seriously until he said that we landed men on the moon. 😂

[@rileybrownai]

you're saying we didn't land on the moon?

[@CATDHD]

Dude

[@Redflowers9]

Come on AI hurry up

[@chrisxavier1848]

The reason it's not going to happen is because consciousness is far more than complex computation. It'll get close, but it won't get there.

[@rileybrownai]

Maybe, maybe not

[@socialenigma4476]

It doesn't need to be conscious to be intelligent. Those are two fundamentally different properties.

[@Pawnsappsee]

For me, it doesn't have to be conscious overall, it just needs to have a good understanding of us as a species and serve its purpose of helping us in every area, then, we're gonna reap many good things.

[@DexieTheSheep]

Beep boop.

[@rileybrownai]

bop

[@DexieTheSheep]

@@rileybrownai [ ☑I am not a robot. ]

[@krox477]

Humanity has to go inward then otherwise we will use all this power for greed and destruction

[@psychocritico5738]

🗿🍷

[@krissnoe500]

Humanity individuals need a purpose ei) a story all their own. Games art challenges and to not fall victim to the "Comparison is the death of Joy." Pattern.

[@rileybrownai]

this is 100% accurate. As a content creator I fall into this trap. It’s not good.

[@matthewkeating-od6rl]

so add blockers imagine data blockers.

[@rileybrownai]

hmmm

[@jonhughes7264]

Really like this content guys, thanks. But I don't think any company is optimising any technology for the purpose of human happiness, now or in the future.

[@Pawnsappsee]

Profit will ever be the goal of companies, but it doesn't mean we'll never get good indirect outcomes from it.

[@michaelwoodby5261]

It's gaHto, not gaYto, goddamn it. You can hear him say it!

[@rileybrownai]

Tomato tomahto

[@RobinCheung]

(consider this moment like I'm Riker grabbing you by the shoulders and shaking you like a baby or whatever the cool kids do now-- --to knock some literal logic back into you, in the place of all top commoner sense, which just. Doesn't. Work. And neither will any of us until you and whoever is meant to, which is NOT everybody and you gotta start practising telling yourself the TRUTH not what will appease people who just it isn't their thing, and forcing them to be involved in YOUR future living or dying is just not at all valuing your gifts that God granted you and hoped you'd set example by seeing they logically aligned synergistically to suggest a super humanly self actializing thing called a CALLING which used to be a thing before it was bombed to sh"t as a direct hit in the war against God

[@paulurban2]

Not a race condition.

[@rileybrownai]

wym

[@thomasr22272]

I'm really tired of those clickbait AI videos, announcing AGI will be here in 18 months or any other specific number is plainly stupid and arrogant. KZbin is full of those types of guys who think they are special and more intelligent than the rest of the other experts, this is just pathetic tbh. The arguments he listed (more money, more parameters, more scientific papers, blabla..) just show there is a trend towards progress in AI, thats it... How can he infer from this that AGI will be here in 18 months? Why not 19 months, or 19 days or 20 years? Well I hope he will make some money views and money out of this at least...

[@clarkd1955]

Would anyone say that Wikipedia is “smart”? Useful certainly but probably not smart. If you put 10 times more data and maybe more topics into Wikipedia, would it be smarter? More useful maybe but not smarter. So more useful doesn’t equate to smarter, at least for information stores. I visualize an LLM as an English speaking person getting some sequence of Chinese characters on a paper through an input tray. The person puts the paper in his machine and turns the crank. Out comes another paper with Chinese characters on it and he puts this paper into the output tray. A Chinese person reads this text and is amazed by it’s wisdom and so asks the person what it means. The person says they have no clue as they don’t understand Chinese. The LLM knows there are lots of numbers in huge arrays. It knows how to turn the crank. It doesn’t understand language at all and it has no access to the data that trained the model. It also has no agency so it isn’t even an it. It has no desires, goals or any capability at all other than it knows how to use the crank to generate output if it is given some input. Does this sound like a “non human intelligence”? I am sure that some versions of LLM’s will be useful but I see no way a model built like this could ever have agency, desires or be any risk to humans. I can see no path, no matter the size of the training set, that would make any reasonable person believe that LLM’s could have intelligence like a human. Useful, maybe but not human or dangerous.

[@grzegorzowczarek3016]

I think you oversimplified how they work. Generally yes, so called "Chinese room" is good thought experiment. But LLM's are NOT just representation of data, or algorithm for making words therefore are not simply not Chinese room. Sabine Hossenfelder have a great video about that, and she changed her mind about AI and understanding. I find her argument compelling. Fundamentally I think it's similar to problem of organic chemistry, and "vital force". People at a time did not fully understand a problem so they made up some special force, and called it a day. For now we see that AI's have limitations so we add some special feature to human brain, while there is no reason to think that brain is not just that: huge neural network changing in accordance to input data. We also do not have access to our training data, what is real is just a structure that arose because of that data. I also do not understand why "reasoning" or "understanding" or "self awareness" are necessary to be dangerous. I find it very much philosophic discussions about semantic.

[@manslaughterinc.9135]

It's pronounced Gato not Gayto.

[@rileybrownai]

oops

[@Doctorstix]

GAYTOE😹😹😹😹😹

[@jonathanlindsey8864]

Thank you

[@rileybrownai]

sorry man.... my b

[@jonathanlindsey8864]

@@rileybrownai lol all in good fun. Thanks for having David on and putting up with him, he's a great futurist, but can be a lot. I truly hope that he has a lasting impact on committees or are some regulatory force for the embodiment of heuristic functions. People like you who help spread his ideas make that possible.

[@j.d.4697]

UBI

[@MiguelMejia-ru4lm]

I remind everyone years ago AGI had came out and lost they are still denied

[@rileybrownai]

they'll see

[@user-rr1dn1no2l]

We will be rewarded with UBI if we lock 🔒 in to the Metaverse and give 🧠 revenue about Horizon Worlds service and dream 😴 time ads 🫥