13 months of AI lessons in 40 minutes - Sottotitoli bilingue
I've been researching AI for 13 months, interviewed dozens of experts, spent thousands of hours learning about AI, and the following of 77,000 subscribers.
I'm going to compress 13 months of brutal AI truths and lessons into this one video.
The lesson is that the smart gets smarter.
And this is a concept from Balaji Srinivasan.
and it's basically the AI version of the rich cat richer right everybody knows that a rich person has extra money
So they invested and that goes even more well per person lives paycheck to paycheck For AI,
the smart gets smarter,
so the people on the cutting edge get even more on the cutting edge,
get even higher advantage, while the people who are behind fall even higher behind.
And that is because they are not using AI.
You cannot really compete with somebody who has teams of AI agents working for them,
who has automations in place, who's using the greatest latest language.
Honestly, this is going to be like in the future, right?
Right now we don't even see this that much,
but in the future the smart will really get smarter to the point of having super intelligent AI at your fingertips.
Imagine-7, right?
If somebody has access to GPT-7 and knows how to use it,
that person cannot even be in the same realm as somebody who doesn't like
it's not even fair right this is why 13 months ago I dropped everything I stopped
uploading on my gaming channel and I went all in AI because I realized that
getting into this field and learning about it now will have a huge advantage in
10 years where other people wake up to it and see the field potential so if
you think you're late don't worry or still very early.
The idea number two is that technology always keeps on improving and honestly this is like the underlying idea behind all of AI.
So the people who don't take AI seriously they don't have this belief or they didn't realize this truth that technology like betting
against the You don't want to do that, like that's a horrible bet.
It's much wiser to realize that technology is always improving,
meaning AI will always get smarter and smarter,
more competent,
more useful,
which eventually will lead to AGI and even later to ASI,
which artificial super intelligence, where the AI is just smarter than all of humanity combined times a billion.
That might seem crazy if you're new to AI, but honestly you have to start with the fundamental belief that technology always keeps on improving.
So it might take 10 years, it might take 50 years.
What matters is that if we really keep improving compute power,
which is one exponential curve,
the Moore's Law,
if we keep improving the number of parameters in models,
or also the efficiency of the language models,
for example, LAMA-37B, which is this year's open source best model, is as good as LAMA-270B.
So a model that is basically 10 times bigger from last year, and that's just pure efficiency.
And there are many of these exponential curves happening in AI.
Overall like betting against the AI feels perhaps one of the worst things you can do because technology always keeps on improving
People tried to ban the printing press when it came out and that was a horrible decision.
I think it was like the wasn't called Byzantine Empire or something like that or no I think it was the Ottoman Empire actually.
They tried to bend the printing press and they completely fell behind because of this decision.
So understand that technology always keeps on improving and this alone should make you take AI seriously.
Now most people are afraid that AI will replace jobs which certainly will happen however AI will replace tasks not a job.
Before we see like AI doing things.
It will be replacing small parts of jobs, which also called tasks, meaning the stuff.
Everybody has a job that has things that could be automated.
AI will replace the things that are the most easy to automate.
It will not do the most complex, nuanced, creative stuff.
It will look at the stuff that's repetitive, that's boring.
that will be automated by AI.
So of being like, oh my god, AI will take my job.
No, it will take parts of your job, which will then hopefully allow you to focus on higher leverage activities.
So this is why I don't see AI as bad because I understand that first it will take parts of the most repetitive.
Like everybody has stuff they hate about their business, about their job.
It's like when they do the studies it's 80%
of people either dislike or straight up hate their job But suddenly when AI is you know or robotics is advancing people are like oh my god
I want to keep my job that I hate
Why not just automate the parts of your job that you dislike you know the most And focus on the high leverage value-producing activities.
This is exactly what AI can do and why I do not think and by the way I learned this from Sam Altman,
which is super AI will not replace jobs, at least not in the near future, it will replace tasks.
And somebody gets fired,
it is simply because they've been doing that tasks, that was easy to replace, not because AI is taking over entire jobs.
One thing I realized recently is that lawyers are basically large language models.
Let me explain.
So, what is JGPT right?
Well, most people use JGPT to get advice.
It's just predicting the next token, giving you the most.
Output based on your query.
What is a lawyer doing or a tax advisor, right?
Well, they are giving you advice, but instead of paying 20 bucks a month or actually check GPT is now
free They charge you like $1,000 an hour or something ridiculous, right?
So So this is where people can really start to see the potential of AI,
nobody really likes to deal with lawyers let's be honest, nobody really likes to do their taxes.
What if GPT-6 could do that for you and it would be better than even the best tax advisor,
also it would be much cheaper obviously and faster.
this is really insane that if you think about it, there are professions that are basically glorified large language models.
Now, I think lawyers will not be replaced immediately because they are protected by let's say regulations, right?
And it's good to have like a human with empathy when defending somebody.
So there are definitely some aspects to human lawyers that are beneficial,
but But if you think about like the parts or like dealing with certain positions that you hate or at the airport,
right, like the incoming suitcases at TSHX, that is just image recognition.
Even the best humans cannot compete with some image model, image classifier.
Millions images of knives or guns or whatever and can just spot them in an instant, right?
A human is sitting there like, you know, bored at the job, maybe spot some war bottle or whatever, but it cannot compete with AI.
And I think this is the main problem is that AI will do the jobs and the positions that don't really add value to society.
automate the things that people hate about life and about dealing with certain businesses or positions.
When you get into the AI field, you will the code, the source code of it, will not be released.
The and the data sets will not be available.
I believe this is the greatest danger of AI.
I seriously think that we have to make sure that AGI is open source and decentralized.
That way, no single entity, no hidden group of five people has control of what we think.
Because this is another thing.
People realize that when you use chat GPT, it is not, you are not fine tuning it.
It is fine tuning you.
Every single prompt, especially the ones that are like on your beliefs or something you are not sure about, it is fine.
making you more and more aligned with the biases built into that large language model,
which in case of shared GPT is obviously OpenAI, which located in San Francisco.
So, if you don't want the typical tech San Francisco beliefs, You have to be careful what you listen to, right?
Imagine Obviously,
JGBT is very limited compared to AGI But once AGI is here,
once we have really something smarter than a human in all aspects You have to make sure that is open source.
You want to make sure that has your best interest in mind Like how do you know like even if that is in a robot?
Let's say how do you know that a robot in your house has your best interest?
Like I would never get a robot a human robot into my house.
That is not okay open source.
Like, I want to see what it's programmed to do.
I want to see if there isn't some backdoor or something sketchy in the code, right?
This is why I seriously believe that the greatest danger to AI is AGI not being open source.
And I think we have to do everything in our power to make sure that AGI is created open source.
One thing that I'm confident in.
unrestricted open source models will actually overtake the cutting edge best closed source models.
Now, why do I think this?
Why do I think that me and you in six to 12 months will be using a worse model,
a less intelligent model than GBD5 or whatever is the best model at that time.
Why do I think we will use something down Basically, it is because it will be more useful.
If you have an unrestricted model that answers everything,
it will be more useful to use cases where chat GPD is hesitant to give you the answer kind of beats around the bush
or just straight up refuses, right?
So, let's see if you are writing a novel and there is some bloodies.
JGVD will just not help you there, right?
Or let's say you are doing something where Example, right?
I a video that was about unrestricted models,
funnily enough,
and Claude Free actually refused help me generate titles for that video because it was like,
I don't believe the ethics of a,
you know,
unaligned AI,
could be dangerous potentially dangerous like that's crazy imagine for your
specific use case where it might be something with like culture
religion your beliefs whatever it doesn't matter you want to understand the model to answer everything,
you don't want it to you don't want to even prompt engineering to like try to like imagine if it's a movie where someone steals a car like why not just say the thing you need
help with and then get the answer so I think I would happily sacrifice five ten percent of IQ or performance for a model that actually answers everything and that's why I think like sitting Or
in 12 months,
probably 12 months,
let's be honest,
we will be using models that run locally,
or if you have a bad computer,
you can run it obviously in cloud,
but a model that just answers everything instead of super restricted cloud free or GPT-4,
which, I mean, everybody who used it, knows how just annoying.
It's like,
permissive the guidelines are,
and once you use a truly landscape model, it just feels liberating, you're gonna ask it anything and it will help you with anything.
One way to think about the potentials of AI is to compare it to other technologies.
So about working, right?
If you need to get somewhere that is far away and you need to walk here, work there, and it takes you four hours.
Well, if you have a bicycle, it might take you just one hour.
That is the power of technology.
At the end of the day, AI is just new technology and it gives you an immense advantage over people that don't have it.
So, if somebody wants to transport through the And they don't have a car, they don't have anything, just walking, that's like super low leverage.
If somebody has a bike, that person instantly gains insane advantage.
And obviously after a bike comes a car, after a car comes a train, there comes an airplane, there a rocket.
There are many levels to this, right?
So AI is simply a technology that gives you leverage.
It helps you do more,
achieve with less and honestly right now I would say we're probably at like the bicycle slash car stage like we still haven't seen most benefits of AI,
that's why most people are ignoring it, they're looking at the current look at the future.
This is a mistake,
you have to look into the future, you have to look where will this technology be 5, 10 years from now and prepare for that.
Okay, I wouldn't be in this industry.
I know it's going to be huge, 10 years from now, even 5 years from now, it's going to be massive.
I wouldn't be there now.
People are looking at the current,
a lot of language models or agents,
and looking at what they can do now, be like, oh my God, the agent failed to do this.
The agents cannot do that.
Agents are useless.
Or AI is over high.
maybe today but certainly not in three,
four, five years from now so that's why you really have to compare it to other technologies such as walking and riding bicycle.
If you don't have to bicycle like imagine 200 years ago or whatever somebody comes to you and like hey I have this bicycle I know you've been
walking to your job for two hours every day in the morning coming back for two hours in the evening.
I got the solution it's a bike.
Now it just takes 20 minutes or 30 minutes for you to get there.
Nobody would be like ah keep that for you.
I my long commute.
Same with AI.
If AI can automate the part of your boring job as I said earlier nobody will kind of resist that right.
Once you see the powers large language walls of AI,
once you see what it can really do, you're not gonna go back and be like, okay, yeah, I'm not gonna use it.
I'm gonna ignore all this amazing technology.
So that's why the working and bicycle example is really powerful.
Let's say you can start a business, but you don't have to hire somebody.
You from the get go, have 50 employees underneath.
That is the promise of AI agents really like before that starting a business was
inaccessible to many people right let's say 50 years ago 100 years ago where you
had to actually hire people because there was no like computer technology
or whatever now AI agents will allow everybody to have
50 employees underneath you doing tasks that you don't want to do or a task that you are not good at doing them for you.
And think this vision is actually so empowering like imagine actually having 50 employees for pennies on the door.
I mean,
if you know anything of API costs,
it is that It's you can like for dollars you can get millions of tokens meaning and obviously that costs are going down
So when GPT for all recently released it they announced it that it was like two to three times cheaper than GPT for turbo
Which was two times cheaper than the original GPT for so the costs are only going down
So in the future all of us will have AI agents that are running
essentially for free doing tasks for us 24 7 wireless
sleeping There's gonna be an agent that checks or email summarizes the most useful ones or whatever and it gives you a
summary Or there's gonna be an agent that searches the new 20 minutes, right?
It you the summary of the most crucial AI news or other news, if you're like watching football or whatever, it doesn't matter, right?
But you're gonna have AI agents doing stuff on autopilot 24-7 for you, for pennies on the door, basically free.
And like the power you will have.
Like that's why I believe that Naval Ravikant is right when he says there will be a billion companies.
There'll be...
Literally billions of companies where it's like either one person,
two people,
just doing something super specialized,
super that they can be the best at and everyone will have,
you know,
30, 40, 50 agents underneath them, working them, giving them more leverage and helping them get done in a week that right now would take a year.
So the reason why I think personally I was so serious about the AI revolution is that I missed the previous two revolutions,
which were,
obviously, the internet wave, the dot-com bubble of 2000 and 2001, and then the social media revolution of 2007 to 10, I guess.
You know,
obviously, I was like barely even born for the internet revolution, which really happened in the 90s, like,
And then for the social media age, I was like super young, basically a child, so I couldn't do anything about that.
This is like the first technological revolution that I can actually be a part of.
That's why, when I realized that...
I, it wasn't really like a hard decision, obviously it's a hard decision to completely
drop everything,
especially a successful channel, so a successful YouTube channel would be a complete dream for most people to completely drop that and go from fresh.
start everything,
no association,
no cross promotion,
new name,
I didn't use my faces in my gaming channel,
so literally zero people cross over,
I had to drop everything,
completely new industry,
a new niche,
a new video format,
like man,
obviously it's a huge risk on paper,
but like for me the decision was kind of obvious,
like the moment I realized this is my chance to be part of a once-in-a-lifetime technology It was kind of obvious.
It like kind of set in stone like it was only a of time Obviously,
I took a few months to kind of convince myself into it,
but deep down I know I knew that I was gonna make the switch and Honestly,
like right now obviously 13 months later,
I can say that it was the right move but all along I knew And this is exactly the type of decision that most people would not make,
therefore most people don't have the benefits.
If somebody is making 20k a month doing YouTube, like that's insane amount of money for the vast majority of people.
There's no way they would...
get that to basically zero,
completely kill the channel, which, by the way, was averaging 1.2 million views a video, stop uploading and go to something for scratch.
But again, as Warren Buffett says, it's not about how hard you roll, it's about the boat you're in.
Alright, so you're probably watching this and thinking like, David, but why should people care about AI?
It's gonna change everything basically.
That is the short answer.
The long answer is that it will let them do the things they always wanted to do with less effort, less money, less time.
It will potentially cure all diseases.
Help us achieve 200, 300 years average lifespan, but more on that later, it will change everything we know about the society.
Like things that matter right now, people are working for money for desires, pleasures, those things might not even be relevant.
AI even potentially...
solve like warfare, right?
If there was a super intelligent entity that could show us like,
hey, the guys on the other side are not as bad as the media makes it out to be, then we would
probably not send our kids, our sons, to those wars.
and there would potentially be world peace for the first time.
Obviously those are super long-term things, but like, why should people care about AI?
Well, because what else would they care about?
Like, it's the biggest thing happening in the world right now.
We're literally building a new species.
There's a new kind of entity being,
200 years ago, let's say, I keep saying this, but whatever, for not all of history, basically, it was the idea that God created us.
Well, it turns out we are actually creating God, right?
We're creating a super intelligent being that's going to be smarter than all of humanity combined.
Times trillion, I mean, it's really incomprehensible.
Once AI starts improving on itself, all bets are off.
I it's kind of impossible to predict after that point.
Like, after the singularity happens, you don't know what's going to happen.
So everything I'm honestly talking about in this video is before the singularity.
Because after the singularity,
it's like,
it's like being consumed by a black,
call while a supernova is happening,
I don't know how to describe it,
nobody really does, like AI starts improving on itself and once it is a super intelligent, better, smarter all humans combined in history.
anything can happen.
So one of the things I hear from beginners especially is that it's easy to get overwhelmed when looking and reading all the AI news right and there is a simple hack for this
that lets you avoid all shiny object syndrome and that is simply ignore everything for seven days.
If something in the AI space is being talked about more than seven days after it's released
It probably has some significance the truth is that
Most products most tools
Most ideas are completely forgotten about like two weeks after a release like let's be honest The air spaces like as announcements every other day.
So if something is Talked about 7 or more days after this release, it probably matters.
So, in order to avoid overwhelm and shiny object syndrome, just like, don't pay attention
to things and only look at stuff that people have been talking about for a week or longer.
Because that, that is probably the thing that you should focus on.
And this simple hack lets you feel less overwhelmed and more focused while still not sacrificing being on the cutting edge.
So I already touched on this briefly but living forever.
Will AI allow us to live forever?
Well, first of all, that is far into the future.
Let's start with a middle step, right?
What is between our current lifespan of, let's say, 80 years and living forever?
that might be living to 200 years, right?
And when you're still living forever, I don't know why maybe it's like religious reasons.
People just have a weird reaction when you say that living forever might be possible or that somebody could live forever or if you ask them,
do you want to live forever?
They're no, I don't know live forever, but they've never really thought about it.
like if you ask two questions, you realize they're just repeating something they heard.
So what's the middle step?
Well, that is just living to like 300 years, right?
If you ask anybody,
if you had the chance to live to 300 years old with perfect health, let's say your health from winter 18 or 20.
Perfect health, maybe if you 300 years.
Would you do it or would you die at 80 with miserable health and pains everywhere?
Obviously 99% of people probably hopefully 100% would say yes.
I would live to 300 in health.
Well, that is the middle step between living to forever.
So, don't even worry about living forever.
For now, let's just focus on extending our lifespans and improving our health.
And the, okay, back to the question, right?
Will AI let us live forever?
Well, the answer is not if the answer is when.
Because again, all you have to go back to the fundamental belief of this video and of AI.
Technology keeps on improving, which means that We will have nanobots.
We will have proteins that can help us cure any diseases.
We have new technologies that can target specific parts of your body to kill cancer cells.
There be so much stuff, it's even impossible to predict.
Again, it's not like you, I can't tell you when that happens, obviously, but I can tell you that eventually it will happen.
Even if it takes 100 years,
the point is that eventually those things will be solved and once aging no longer exists, the concept of living forever becomes possible.
Obviously, the U.S.
still die from an accident.
some freak somebody wanting to kill you,
but like dying from aging that
Could very feasibly very likely be solved in the next 50 years one of the biggest
mistakes I see beginners make especially when it comes to agents is Having unrealistic expectations.
So obviously everybody wants to automate half of that their work, they would automate something that takes four hours, right?
That is the incorrect approach.
What you want to do is you want to find the smallest possible things in your day to day,
or even if it happens once a week or once a month, it doesn't matter.
Just small tasks that are easily And start there because if you do that you will get the benefits of compounding AI automation,
so Automating something that takes two minutes every week like two minutes a week that is not inspiring that is not you know motivating
But that adds up There's 50 weeks in a year or 52, right?
So two minutes times 52, that's like one or four.
That's one or four minutes per year.
And that's just one small automation.
Again, that's the goal, the goal is to compound.
If you add a two minute safe here, five minutes safe there, you know, three minutes safe there.
And you have like more and more of these small little time safes like with agents and, you know, just AI in general.
Over time,
you'll find yourself having way more free time than somebody who doesn't have it and it's gonna be the things that are the most boring and repetitive and you're
You're gonna find yourself doing the things that actually bring you joy make you fulfilled and create value in the world while
All of these like small pesky tasks will be done by AI agents.
So you have to make sure that You get the compounding benefits of agents,
you by analyzing all the tasks you do and choosing one, you have to choose one to get to building, honestly that is the secret.
There is a lot of friction at the start, but you have to build your first agent, your two agents.
It becomes infinitely easier and this is the same with everything like going to the gym for the first time scary
Intimidating but when you go to the gym for the 100th time, it's like literally for effortless.
It's like breathing hair So with AI agents just do the same you have to start the compounding process same with investing with investing with this invest early when you're young and
Keep investing so that it compounds with agents
You have to start with something small something easy to automate and build on top of
that they are two minutes there and you'll be shocked in two years just how much time you have.
Now AI safety.
The term AI safety is one of the first terms you'll hear here in AI let's be honest and it kind of sounds good.
It's kind of like the term federal reserve.
It's supposed to sound Interesting and relevant, but as we know with the Federal Reserve, it is not a federal institution and it doesn't hold.
AI safety, like it's one of those things that everybody wants to support.
Oh yeah, of course not for AI safety, but wants to really dig into it and realize what it means.
You realize that it's kind of an unfounded movement.
Basically, it's people who think that AI will just one day decide to kill us all, choosing to put regulations on it today.
So if you look at shared GPT or any language model, literally any AI model that exists today, that are precisely zero worrying.
Things about it.
It is literally predicting the next token and that's exactly it.
That's it.
That's it.
There is no like developing secret agenda of Escaping the computer or there is no like,
you know,
secret communication language that it tries to encode or install a when you didn't do it Agents like people who don't understand agents,
they just assume that whatever agent goes rogue and suddenly uninstalls all progress on your computer,
first of all, that impossible because if you don't give it, that rules will not do it.
Second of all, it is breaking this next token, so it is following the rules.
It's literally the architecture doesn't prevent it, doesn't let it do that.
If you give it instructions, that is the first tokens, right?
It is predicting next token based on that,
so if you give it instructions to search the web,
there is zero percent likelihood that the next tokens are going to be
uninstalled to computers and install a key logo and still all the passwords.
So AI safety.
Completely unfounded and the problem with that is that if you regulate something that is the potential to save lives
and Free people from the most slavish jobs
I mean most of you watching this probably have a pretty decent job
But there are people working in factories who do this all day just like moving something from a you know assembly line into a box
Literally they do this all day eight hours straight absolutely mind
-numbing like man insanely bad job AI and robotics has the potential to free us from that and let us focus
on things that actually give us value give us joy and fulfill us.
So AI safety is a huge issue because it is a movement based on nothing.
There is nothing worrying about the current state of AI or even the next version of LLMs that
could potentially mean ban all AI and destroy all the potential wonders.
So why I think ASAC is a massively overhyped problem and most beginners don't understand this.
So let's say you suddenly lose access to the internet, right?
You would be screwed.
I you couldn't really do anything in this world.
Obviously, you can go live in the woods, but let's be honest, nobody wants to do that.
If the government decided,
hey, you are no longer allowed to use computers, the internet, and there is no way you have to follow the law.
You would be completely screwed.
You couldn't compete with other people in business.
You couldn't talk to your friends online.
You would just be completely gone out of this system.
Now with AI, that's gonna be much, much worse if you don't have access to AI.
Obviously today, it's, again, this is the issue, right?
People who get AI today are like, I could live without chat GPT, I could live without LLMs, but that's not the point.
The point is that when we have super intelligent,
like 200 IQ, models that literally can give you the best possible advice you might get and any single area of life.
If you don't have access to that and somebody else does, there is no way you are ever going to catch up to that person.
Like person will destroy you in any aspect of life.
sport, whatever people compete and, you know, compare it themselves on.
So this is exactly why it has to be decentralized.
It needs to be permissionless.
It needs to be peer to peer.
There cannot be some central entity that decides,
hey, you no longer have access to AGI, because if that's the case, you're done, like you're completely screwed.
So not having access to AI,
And five years from now will be just as bad, if not worse, as not having access to the internet today.
Oh, come I'm not going to bounce.
There's a question whether you should learn programming if you want to get into AI.
Some people think that programming will be replaced, which honestly might be the case.
We might do all programming in straight English.
But right now, today, if you want to get into AI, knowing the basics of programming and Python is a huge advantage.
I mean, my belief is that you're never going to get worse by learning a new skill, right?
Even if you learn programming and in five years, we don't use any programming languages, we just do so.
So what?
You understood the logic and you know how it works below the surface, which is already an over people who don't know how it works.
So, if you want to get into AI, should you learn programming?
My answer is, of course, like, why not?
well, agents, right the best AI frameworks all involve Python.
And obviously, you don't have to be an expert.
You just have to know the basics, you have to understand data types for loops.
If else statements, like the absolute basics of programming, then the syntax of Python, if If you don't you will not be intimidated by agents.
There so many people that believe in agents want to get into building agents,
but when they watch a tutorial and somebody opens up VS code and starts writing some code,
they're immediately like, intimidated basically and stop watching just because there's some code in there.
involved.
Why?
Why not spend a few hours learning the basics of programming which honestly is not that hard.
It's like basically just logic at the end of the day.
If you learned it that you can build agents,
you can code your own Obviously use AI to help you with that because why would you do everything yourself?
But like there is no disadvantage to learning programming.
Just a new skill in your arsenal always makes you better Okay, so let me debunk this once and for all.
AI doesn't have emotions.
It's not self-aware This was like so annoying to see especially when Claudefully released that so many people who don't know how LLM's work
They don't know how next token prediction actually works.
We're like Cloud-free is self-aware.
It this,
you know,
when the researchers asked it,
gave it the needle in a haystack problem, which is basically just putting a sentence into a long piece of text.
It was like, oh, this sentence seems out of place.
Seems like you might be testing me whether I can remember it.
Again.
it is predicting the next token so it just makes it it just like yeah the model is good obviously
so the answer was impressive the answer was the most likely thing that would be happening if you
gave a human 20 sentences and in the middle was some like vulgar joke they would be like what is
that like why is that there do you just trust me that i notice it Same with the LLM,
it's predicting the next token,
it's not self-aware, it doesn't have any feelings or opinions about itself, it simply is a large language model, it's like...
a neural network, it's a bunch of math, if you boil it down to the most fundamental level, it's math.
So if somebody wants to ban AI, they want to ban math, like it's matrix multiplication.
You have matrices and they're multiplied against each other.
And if that's done on a big enough scale, you get intelligence out of it.
So, are LLM's self-aware?
No.
Are LLM's emotional?
No.
In the future, AI might become emotional, might conscious.
but right now there are no signs of it and the current elements are nowhere near being self aware.
So you see anything about that please go to like neural network basics,
machine learning basics, and learn how large language models and next-to-comprection work.
you will know like how ridiculous these claims are.
UBI, the good old concept of universal basic income that everybody is getting a check every month and they don't have to work.
A lot of people believe that this is the future because once AI replaces more and more jobs,
you know,
once it goes from replacing tasks into full jobs,
Those people will have to survive somehow, so, like a very popular belief in the AI community is that UBI is the answer.
And a lot of people don't even bother to think for different solutions,
they just say like, oh, UBI is the answer because that's all they know.
My problem with UBI is that you are completely and utterly reliant on the government.
If for whatever reason the government decides, oh, you didn't follow law number 8731, sorry, no check for you, suddenly what?
Most people are living paycheck to paycheck,
so if you don't get your check in the mail, how do you buy food, how do you pay the bills?
Like, I think the concept of UBI has massive problems with it.
I'm not saying it's not like part of a solution or I'm not saying it's not going to be implemented.
Obviously some countries will try to implement it.
I'm just saying that we should be very very careful with UBI.
because it just gives complete control over the government and kind of,
you know, kind of reminds me of that Klaus Schwab, like, you'll know, you'll own nothing and be happy.
Like, personally, I don't want to do that.
And I'm not interested in being super 100% reliant on the government.
So if UBI ever exists, like, I don't care.
I'm making sure I Some bureaucrat or some politician is not gonna depend whether my family survives.
So UBI, a dangerous concept that we have to be wary of.
So obviously everyone wants to know, what can I do to not get left behind?
This is one of the main things people and myself included get into AI.
So, okay, what can I break it down into?
started by learning about AI.
You set by 30 minutes a day,
every single day and just research about AI and don't be lazy,
don't go to YouTube and just,
you know,
get to recommend Look at some research papers,
look into different news,
maybe look into some Twitter accounts that are credible and reputable, maybe take a course by deeplearning.ai, which is by Andrew Aang, amazing resources.
You can do so many things to not get left behind, it's more about just doing them, right?
You don't have to do everything but anything is better than nothing.
So, learning about AI is the main thing, is the beginning.
You to learn about it constantly to really realize its powers and why you should care about it.
Next thing is actually building something,
whether that's an AI agent,
whether that's just using mid-journey and making some images,
actually achieving something with it, or going to change GBT and creating your own custom GBT for something.
That alone is like huge.
The next step is Try as many AI tools as possible.
You're not going to use all of them But try them and actually try to pay the subscription because so many people are cheap
They don't want to spend $10 $20 trying some AI tool that might save them hours every month
You don't know most tools have the best options behind the paywall.
If you're not going to pay those 20 bucks you're never going to see the full potential of that tool that might completely change
your business or productivity.
So try as many tools as possible and see which ones you stick and you think are worth the money.
That's another thing to not If this is one of the easiest things,
applying peer pressure in a positive way,
everybody knows peer pressure in like negative scenarios,
for example, somebody forcing you to drink, somebody forcing you to smoke, but peer pressure works just as well for positive things.
So, if you're surrounded with people, all of which who are serious about AI, you're not going to be the only one who has a free version of JGPT.
No, you are going to
fully focused,
fully committed and you're going to be on the cutting edge and one thing is to share if you are surrounded by people who take AI seriously you're not going to be left behind
and this is why I created a community for AI people the new society was created for this exact
purpose to combine all of the people who take AI seriously all the people who are actually building the future.
We're on the cutting edge of this technology, put them into one place and just let the brainstorming happen, let the networking effect take place.
So, if that's not interesting to you and if you want to join my community and get exclusive access to myself, we do free calls a week, every week, then check it out, it's the first
link in the description.
Lingua di traduzione
Seleziona la lingua di traduzione
Sblocca altre funzionalità
Installa l'estensione Trancy per sbloccare altre funzionalità, tra cui sottotitoli AI, definizioni di parole AI, analisi grammaticale AI, parlato AI, ecc.
Compatibile con le principali piattaforme video
Trancy non fornisce solo supporto per sottotitoli bilingue su piattaforme come YouTube, Netflix, Udemy, Disney+, TED, edX, Kehan, Coursera, ma offre anche traduzione di parole/frasi AI, traduzione immersiva a tutto testo e altre funzionalità per pagine web regolari. È un vero assistente per l'apprendimento delle lingue tutto in uno.
Tutti i browser delle piattaforme
Trancy supporta tutti i browser delle piattaforme, inclusa l'estensione per il browser iOS Safari.
Modalità di visualizzazione multiple
Supporta modalità teatro, lettura, mista e altre modalità di visualizzazione per un'esperienza bilingue completa.
Modalità di pratica multiple
Supporta dettatura di frasi, valutazione orale, scelta multipla, dettatura e altre modalità di pratica.
Sommario video AI
Utilizza OpenAI per riassumere i video e comprendere rapidamente i contenuti chiave.
Sottotitoli AI
Genera sottotitoli AI precisi e veloci per YouTube in soli 3-5 minuti.
Definizioni di parole AI
Tocca le parole nei sottotitoli per cercarne le definizioni, con definizioni alimentate da AI.
Analisi grammaticale AI
Analizza la grammatica delle frasi per comprendere rapidamente il significato delle frasi e padroneggiare punti grammaticali difficili.
Altre funzionalità web
Oltre ai sottotitoli video bilingue, Trancy fornisce anche traduzione di parole e traduzione a tutto testo per pagine web.