Latest / Bad Decisions Podcast / $122 BILLION to make CHATGPT the AI Super App 👀
Transcript
- 0:00AI is massively impacting the job market and you should know
- 0:04why. Open AI raises $122 billion to
- 0:10make the AI Super app. And we'll be talking about an AI
- 0:14tool that is actually helping 3D artists and game developers.
- 0:18Let's talk about it. What is our beautiful people?
- 0:20Welcome back everybody to another episode of Bad Decision
- 0:24Podcast. Where we bring you the latest
- 0:26and the greatest and the hottest in the world of science,
- 0:29technology, AI and business every Monday, every Wednesday
- 0:35and every Friday. Today is Wednesday and we are
- 0:38back with another episode. I have to tell you guys that we
- 0:41just crossed 100 podcast episode for Bad Decision.
- 0:46Today is episode 101. Yes, thank you to every single
- 0:51one of you guys who has been with us on this journey.
- 0:54Whether you were there on episode 1, some of you are
- 0:57there. It was just two of us.
- 1:00Episode 1 remember I believe we had about 100 views on the first
- 1:05episode, or 250. Now we have, of course, more
- 1:08than that. Yeah.
- 1:09But the day in the beginning about 20.
- 1:11To 50, half of which were our family and friends supporting us
- 1:15in the comments. But the fact that you guys have
- 1:17supported us along the way means the world to us.
- 1:20Of course, this podcast has shaped itself in different forms
- 1:24over the past few years, and now it is in its ultimate form, this
- 1:29new format of the show that we're enjoying very much.
- 1:32And we receive your comments. You guys seem to be enjoying it
- 1:35as well. Some of you enjoy the new
- 1:37sections and the intellectual conversations.
- 1:39Others are benefiting from the live demos, but the reason why
- 1:43I'm mentioning this one, of course, is to show how grateful
- 1:47we are to you. And 2nd, the fact that we do
- 1:50read your comments. So if you do have feedback for
- 1:52how we can improve the show, the kinds of topics that you guys
- 1:55want to see, do let us know in the comments because they do
- 1:58come up every episode. I mean we have 3 episodes a week
- 2:01so we have plenty of time to cover everything.
- 2:05I love the comments that mentions Team Farhad.
- 2:07Keep them coming. Well, we have a lot of Team
- 2:09Farhad. One specific comment on YouTube
- 2:12that I responded that said hey, the guy on the left should let
- 2:16the guy on the right. There was there was a comment on
- 2:20Spotify. That's really.
- 2:21Good points he wants to make and he seems to be getting
- 2:24interrupted so. There was a comment on Spotify
- 2:27explicitly mentioned the name of Pharaohs.
- 2:30He mentioned that Pharaohs should stop cutting Pharaohs
- 2:33off. Fat hods?
- 2:35How many far Hods? Ideas.
- 2:39Fat hods, plural. There are so many fans here.
- 2:42I can make multiple agents of myself.
- 2:45We have very important topics to discuss today.
- 2:49This is going to be one of the more heavy topics that we
- 2:54discuss, so I do urge everyone to prepare themselves.
- 2:58It's going to be a pretty decent but heavy conversation.
- 3:02So get ready, get your popcorn, your coffee, your can of Coke,
- 3:06Pratt, I don't know what. What do people want to drink?
- 3:08Pepsi. They also drink Pepsi.
- 3:10Pepsi. OK.
- 3:10Pepsi, Coke. Whatever.
- 3:12Team you're on. Team Coke.
- 3:13Team Pepsi. Just get your drink, get your
- 3:15snack. We're going to get right into
- 3:17it. You ready?
- 3:18Let's go Oracle laid off 30,000 people earlier this morning and
- 3:26Block laid off 40% of their workforce earlier last month.
- 3:34And the reason may surprise you for the first time, this is not
- 3:38the top companies attempting to cut costs.
- 3:42As per usual, they're preparing for a new way of working 1 where
- 3:49AI is revolutionizing every aspect of how we work, who we
- 3:55work with, and in what structure.
- 3:58Whether you like it or not, if you're a human who is working
- 4:02today, this news concerns you. Why are we talking about this
- 4:07topic specifically? Because Jack Dorsey and Rolf
- 4:12released an article on X earlier, I believe today, or it
- 4:16might have been yesterday, and it's titled From Hierarchy to
- 4:21Intelligence. It is over 2 million views as we
- 4:24speak and it is growing pretty rapidly.
- 4:27It's a lengthy article. We recommend that you read it
- 4:30for yourself, but we're going to give you a summary and tell you
- 4:34our take of what we believe is actually happening.
- 4:39The reason why I also mentioned Block earlier on laying off 40%
- 4:44of their workforce is because Jack is the Co founder of Block.
- 4:48If you don't know Jack Dorsey, he's also the original founder
- 4:51of Twitter, the same platform we're on right now, which is
- 4:54called X taken over by Elon, was founded initially by Jack.
- 4:58So he is a visionary, he's a builder, he's been in the
- 5:02technology space for decades, and many people follow him in
- 5:07the ideology of how companies and startups should be run
- 5:11because he's had many successful startups.
- 5:14He knows what works and what doesn't work, and he's always
- 5:17been an activist for the reduction of bureaucracy and the
- 5:23increase of speed and velocity in running businesses.
- 5:26And this article is tackling that very specific topic.
- 5:31Before I talk about what the article is stating, I'd like to
- 5:35tell you guys a little bit about the history of humanity because
- 5:38it relates to what the article is referring to.
- 5:42And Farrad loves it every time I talk about the book that I'm
- 5:45reading, Sapiens. So I'm going to I'm going to
- 5:47bring it up. This is a book that covers the
- 5:50entire history of humanity while attempts to in about 500 pages.
- 5:55Whoever loves to understand how we came to be where we are
- 6:00today, I recommend that you read this book.
- 6:04It is very lengthy. It is incredibly heavy, but it
- 6:08will teach you about your game, the game that you're living,
- 6:11which is this life. In the very beginning.
- 6:14It talks about the history of humankind.
- 6:16And if I were to quickly summarize how we came to where
- 6:21we are today, we've existed for millions of years as the
- 6:27different Homo species, Homo Neanderthals, Homo erectus, Homo
- 6:32sapiens, and for the millions of years that we've existed,
- 6:36initially, we organize ourselves in small groups, 1020, sometimes
- 6:43100. But most animals, most of them
- 6:47break apart after they reach a certain number of people or
- 6:51certain number of units in a group.
- 6:54There's only a few animals in the world that can organize
- 6:58themselves in large groups. Ants are one example, bees,
- 7:03bees, humans are another. But we weren't always like that.
- 7:06We would only form groups of 10 to about 100 and then the
- 7:11hierarchy in those groups would break apart because of power
- 7:14struggles, status and so on and so forth.
- 7:18But the reason why about 30,000 years ago, Homo sapiens became
- 7:22the last remaining species of all different versions of Homo
- 7:29is because we managed to somehow have a cognitive revolution that
- 7:33allowed us to form a much more sophisticated language and to be
- 7:38able to collaborate on a much higher level compared to other
- 7:43species. And excuse me, Homo sapiens
- 7:47since 30,000 years ago, where we've seen the extinction of
- 7:51Homo Neanderthals to about two 3000 years ago where we started
- 7:56empires, large empires consisting of hundreds of
- 8:00thousands of people's invention of money and so on and so forth.
- 8:03The reason why we've become incredibly successful and
- 8:07productive is because we managed to form more and more
- 8:10sophisticated hierarchies where we had the low level workers and
- 8:16then a group on top of them, another group on top of them.
- 8:20And you can keep going until we came to a point where we had the
- 8:24industrial revolution. In the industrial revolution, we
- 8:28went from being in tribes and in small groups of farmers to then
- 8:33working as hundreds of thousands of people under 1 mission.
- 8:41So a great example of that in this book is Pejal.
- 8:46Peugeot is a car company. No one is born thinking I'm
- 8:50going to be someone who works for Peugeot.
- 8:52No one is passionate about that since they were born.
- 8:55But we as humans managed to group up together, work in large
- 9:01numbers under a brand called Peugeot.
- 9:04You can destroy all of the Peugeot cars today, but the
- 9:06brand itself will still exist. You understand what I'm saying?
- 9:10Destroy all of the cars. But because that brand, an
- 9:13imaginary, an imagined order is what it refers to in the book,
- 9:17exist as Peugeot and people believe in it, then it still
- 9:21lives on. And so the fact that we managed
- 9:24to establish that in our brains allowed us to work as
- 9:28hierarchies and large and larger groups.
- 9:31I could be in Malaysia and you could be in Dubai and Pratt
- 9:34could be in Canada, but we could still work remotely for Peugeot.
- 9:39Even if you've never met each other, we can still be working
- 9:41under the same company. And that's something that other
- 9:44animals cannot do. You cannot be in three different
- 9:47parts of the world and work online for the betterment of a
- 9:51name called Pejo. This is what sets us humans
- 9:55apart. It's these hierarchies that we
- 9:57build, some most imaginary hierarchies, and for the longest
- 10:03time these hierarchies have allowed us to become the apex
- 10:08predator, to become the most intelligent species.
- 10:12What this article is stating is that we're seeing such a massive
- 10:18shift due to AI that these hierarchies are no longer
- 10:22efficient. There is a new way of working
- 10:26together that can be much faster and much more reliable.
- 10:32What they're referring to is that hierarchies are ancient
- 10:34hacks for human bandwidth limitation.
- 10:39There's potential data loss, low coordination speeds, and
- 10:43bureaucracy. Let's use a company as an
- 10:45example. You have a marketing team, a
- 10:48design team, and HR team and a lot more.
- 10:51I'm just going to use as those three as an example.
- 10:54You have 10 people in each group.
- 10:56Those 10 people will report to a low level manager.
- 11:00That manager will oversee the work and report to a higher
- 11:03level manager. The higher level managers will
- 11:05then report to the senior leadership.
- 11:08During this relay of information and context up and down the
- 11:12chain, you can have potential data loss amongst the managers
- 11:17and the communication channels are relatively slow.
- 11:22Is there a way to improve it? That is what the article is
- 11:25stating. So they believe that using AI,
- 11:29we can compress coordination dramatically for real speed
- 11:34gains by building world models that can act as Central
- 11:39Intelligence while we have human edge workers reporting
- 11:43essentially and communicating through the AI models that can
- 11:46keep track of everything and relaying the information at
- 11:50speeds that the traditional hierarchies wouldn't be able to.
- 11:55And because Jack traditionally and consistently throughout the
- 12:00years has always been about speed and velocity, and knowing
- 12:05that this is one of the most significant traits of a
- 12:07successful startup is making the assumption that this is going to
- 12:11be the new way that companies will work in the future.
- 12:17I'm going to give it to Farah in a moment because I want to know
- 12:19his opinion. But the reason why he's also
- 12:22written this article is partly to the due to the fact that he
- 12:26just announced a 40% of employee layoffs recently.
- 12:31And he did mention that this is not purely because we're trying
- 12:34to cut costs. This Is Us restructuring.
- 12:37And this is perhaps the first explanation that we're seeing
- 12:40from it. Farhad, I want to know your
- 12:43thoughts so far. When I started reading the
- 12:47article, the first thing that he mentioned, which I totally
- 12:50agreed with, was that most of the companies are using AI as a
- 12:56productivity tool. So everybody in the company
- 12:59Start learning AI, we are going to have a marginal improvement.
- 13:03You're going to be 10% faster, 10% more productive, 10% more
- 13:06efficient. What we think we can do is we
- 13:10can restructure the company over AI.
- 13:13So it's not that AI is helping you.
- 13:15Oh, I'm 10% faster. I'm not using the Excel anymore.
- 13:18No, we put AI in the middle as the core.
- 13:21Now everybody works around it. So it's a structural shift.
- 13:26In theory, probably it's going to work.
- 13:30In practice, I see a lot of downside.
- 13:33Do you want me to go through the downsides or you want to
- 13:36continue with the article? I have personal experiences
- 13:39through this. I would love to hear you talk
- 13:42about it. When you mention I've worked in
- 13:44a big company before starting bad decision with Faros, our
- 13:47company. When I started the company, we
- 13:49were 30 people. I was employee number 31.
- 13:52When I left the company, we had around 300 employees.
- 13:55So the company grew 10X. We started with not having a
- 14:00sales team or a sales department or even AC level to having HR
- 14:04department, sales department, marketing department, exactly
- 14:07like this article, exactly what you were mentioning.
- 14:10Yes, there are data lost. Yes, it was hard.
- 14:13When you become bigger, it's hard to do things.
- 14:15You need to get approval. You need to go talk to this.
- 14:17Person lots of bureaucracy. But that creates culture, that
- 14:22creates human interaction. There was a culture of, oh, this
- 14:27team works better with the other team, or this team can stay
- 14:29more. There was a there was a reason
- 14:31for people to stay in the company because of the culture.
- 14:35We didn't have overtime payment. We didn't have high bonuses.
- 14:38A lot of people stayed in the company because they had friends
- 14:41in the company, because they had culture.
- 14:44What Jack is implying to as Aceoi totally understand his
- 14:48point of view. Of course, you have to be fast,
- 14:50you have to be efficient. But when I look at, even when I
- 14:53look at this graph and then I see humans all not connected in
- 14:57a group, but just, you know, pointing towards a single
- 15:00direction, I would see a lack of culture.
- 15:03And then I'll give you another example.
- 15:05Me personally climbed the ladder, climbed the corporate
- 15:10ladder by not only having technical skills.
- 15:14One might say my technical skills were very limited.
- 15:18I had great social skills. Communication.
- 15:20Skills. I could communicate with my
- 15:22COOI, could communicate with my COOI, could communicate very
- 15:25well. I could lead.
- 15:26May I say that initially when you entered the hierarchy at a
- 15:30low level, you managed to communicate your way up the
- 15:33ladder by providing to the managers what you thought would
- 15:37work. And initially you didn't have
- 15:39the technical skills, but because you made them trust you,
- 15:43over time you we built those skills.
- 15:44I I hadn't been an AI system. That would be out on the 1st.
- 15:49Day that would be incredibly firm with the results you would.
- 15:52Be out. No, I, I would be.
- 15:55So we did a test. So we did a week of training and
- 15:58everybody would do a test. They would teach us SQL, they
- 16:00would teach us HTML coding and some business analyst skills.
- 16:05I failed in almost all of them in the first three months.
- 16:09I was one of the worst hires that we had in the company in
- 16:12the first two years. I was the only one who stayed
- 16:15from my batch and I was the highest rank.
- 16:18So there are a lot of nuances when you are working and humans
- 16:24are going to spend 80% of their time working, which probably in
- 16:28the future with AI, they're not going to spend 80% of their time
- 16:30working. So I totally understand his
- 16:32point of view, but I'm very skeptical and very afraid of now
- 16:37who actually does want to work in that company when there is a
- 16:44very limited culture available when I don't have any meaningful
- 16:50relationship with my Co workers. I have to echo what you just
- 17:00said because in my job as an E sports host slash producer.
- 17:06That is even worse, right? You were in.
- 17:07Malaysia. Environment.
- 17:09I remember precisely. I loved my job, but I personally
- 17:14knew first hand at least 5 to 10 people in the company who didn't
- 17:19like their job. They disliked it.
- 17:22Every day was a struggle. They only stayed because of the
- 17:26people. Yeah, Every time they thought
- 17:28about quitting, they would just remember all of the time they
- 17:33spent together during work hours or after work hours, playing
- 17:37games, having pizza nights. It was the culture that made
- 17:41them stay. And I think COVID and the
- 17:44pandemic partly killed that culture.
- 17:47So many people became used to working remotely like robots.
- 17:51So many people became used to just going online on a Zoom call
- 17:55and reporting on what they've done, perhaps even lie to an
- 17:59extent on what they've achieved because you could, since you're
- 18:02not in the office. Would like lying to the core AI
- 18:05system that is there. Exactly.
- 18:07So I think the pandemic conditioned people to work in
- 18:10this way, to work in a network in which you are isolated from
- 18:16your peers and other people. And so a lot of people would be
- 18:20would be they wouldn't feel like this is actually quite alien to
- 18:25them. That's what I'm saying.
- 18:27I agree with your point, but I think COVID already prepared and
- 18:30paved the way for this way of working.
- 18:32And first thing let. Me, let me get to my point.
- 18:36The guy on the left should not disturb the guy on the right for
- 18:39a moment. Here debate moment.
- 18:46Did I? Yes, I'm going to remember it.
- 18:48I'm going to remember it. The.
- 18:52COVID condition people. Yes, COVID conditioned people.
- 18:55I'm going to bring it back. COVID paved the way for this way
- 19:00of working. No.
- 19:02OK, I lost it. You go.
- 19:04What did I want to say? I said you.
- 19:06Better remember. No, but I do remember what I
- 19:09said is COVID was very different.
- 19:11You were working remotely still with people then.
- 19:14It's so systematic when you are submitting your work to an
- 19:19intelligent system that your manager is an intelligent
- 19:22system. It's just one and zeros.
- 19:25It's binary. There is nothing in between.
- 19:28For big corporation, that's great.
- 19:31For startups, for any companies below 100 people, that is going
- 19:35to kill the company immediately because you're not going to have
- 19:37any culture building that company and.
- 19:39Let's go to the beginning. I agree with you.
- 19:40Let's go to the beginning. I think we can all collectively
- 19:43agree that laying off 30,000 people at once is inhumane.
- 19:49There is no way you can beat around the Bush and say that is
- 19:53justified emotionally. If you speak on an individual
- 19:57level as a human being, how can you be OK with 30,000 people
- 20:02losing their job? So I think we can all
- 20:04collectively agree on that, right Pratt?
- 20:06Yep, you Pratt. Do you have a heart?
- 20:09Are you saying yes or no? Yeah.
- 20:11I have a heart. OK.
- 20:12Oh, yes, you have. Are you?
- 20:14Yes, You're not. And one thing to know, we were
- 20:21discussing this with our friend earlier today.
- 20:23You shouldn't even count it as 30,000 people.
- 20:26Those 30,000 people have families they're paying for.
- 20:30And if they have two members of their family they're taking care
- 20:33of, you're technically laying off 90,000 people.
- 20:36In a sense, you're impacting the lives of almost about 100,000
- 20:41people. And this is just one company.
- 20:44So there is no way you can say that as humane.
- 20:47It is inhumane. What is happening is unfair, and
- 20:51there's no way around that. That's the first point I want to
- 20:54make, just to lay the grounds for this.
- 20:56Now I want to talk about a bird's eye view.
- 21:02Say you're looking as a creator of a game.
- 21:07You're playing Civilization or Age of Empires, and as the
- 21:10player or the creator of the game, whichever, you have one or
- 21:15two goals, and the main goal is to win the game.
- 21:17And to win the game, you need progress, you need economic
- 21:20growth, you need social growth, so on and so forth.
- 21:23And in Sims you also need happiness, right?
- 21:25I've never played Sims, but in Sims you have actually have a
- 21:28happiness level for your human happiness level, how they're
- 21:31behaving, how is the society developing, right?
- 21:34So all even that is important. If you take everything into
- 21:37account, starting from economic growth, of course, what they're
- 21:41saying makes sense, right? The reason why I even started by
- 21:45talking about this book is because if you were to look at
- 21:48humankind as a species, not on an individual level, but as a
- 21:51species, the reason we have one is because of all the things
- 21:55we've done in the past. But go and look at the history
- 21:58of humankind. How many humans have died in
- 22:00violent ways for us to have progress, whether it was for the
- 22:04industrial revolution or the scientific revolution?
- 22:07It all sounds really fancy and sounds like progress only.
- 22:11But if you actually go deeper and understand what it took for
- 22:14people to go through these ages, we've lost so many humans in the
- 22:19process. We get to enjoy the benefits.
- 22:22But many humans were unhappy because of the industrial
- 22:25revolution. Many people didn't want to
- 22:27fucking. Let's go even further down,
- 22:30further back. Many people hated the fact that
- 22:33we went from hunters and gatherers to farmers.
- 22:36Imagine a hunter gatherer. Your job was to wake up and go
- 22:41hunt for food. So the landscape that you
- 22:43covered on a daily basis was massive.
- 22:45You didn't have a single home. You lived everywhere.
- 22:48And to us, that sounds like a dreadful existence.
- 22:52But I guarantee you if you told the hunter gatherer, come and
- 22:55live my life, they would spit on me right now in a room and look
- 22:59at a fucking screen all day long.
- 23:01For me, this is comfort. This is safe.
- 23:04I'm used to this. But for him, he's like, I
- 23:06fucking conquered the mountains. I fought with lions and I had
- 23:11many homes and I didn't attach myself to items and I didn't
- 23:15have consumerism. Fuck your safety.
- 23:18That's what he would say. So for hunter gatherers to
- 23:20become farmers, that must have been dreadful.
- 23:25That transition was not safe. And this book covers it in
- 23:28detail. It says farming and the whole
- 23:31fact that we became farmers, the agricultural revolution was a
- 23:35fraud. We thought that we are
- 23:37domesticating wheat and plants, but in fact, wheat and plants
- 23:41domesticated us. Do you know why?
- 23:44Because think about it this way. We started guarding the plants
- 23:49and wheats. We started taking the seeds and
- 23:52planting them in specific areas, and then we couldn't leave the
- 23:55area because we have to protect and make sure it grows.
- 23:58If you are somebody, if you are a wheat, right, you're not a
- 24:02wheat. But I'm just saying pretend like
- 24:04you are. It seems like the human is being
- 24:08domesticated by you because you're changing the behavior of
- 24:11the human so that it grows you. You change the behavior of all
- 24:16humanity. You changed how they lived, how
- 24:20they made it, what they ate, everything so that you grow.
- 24:25And I don't remember the exact numbers, but it is here in this
- 24:28book. Wheat was just grown naturally
- 24:31in specific parts of the world in small quantities, but within
- 24:36just a couple of years, it grew to almost every part of the
- 24:39planet because we fucking took it everywhere.
- 24:42And so the book is stating that perhaps it wasn't us who
- 24:46domesticated the plants, that plants domesticated us.
- 24:48The wheat domesticated us. Why did I even bring this up?
- 24:54We have gone through many cycles of evolution during humanity,
- 25:01during the course of humanity. And if you look at it as a
- 25:05player of this game, but from a bird's eye view, looking at not
- 25:08just one human, in looking at an overall spectrum of what has
- 25:12happened, there has been a lot of pain along the way, but at
- 25:17the cost of certain humans, other humans benefit.
- 25:22And it seems to be happening now.
- 25:24If I swear by this and say if you were to pull back and be
- 25:29able to live the next 3000 years or at least view the next 3000
- 25:34years. You will see the cost of it is
- 25:37being paid by these 30,000 humans here and a lot more over
- 25:41the next few years and perhaps 50 years down the line or 100
- 25:44years down the line, you'll have humans who'll be like, Oh my
- 25:47God, I love my life now because of XY and Z, but they would
- 25:50forget about these 30,000 people that actually had to suffer so
- 25:56that they have a better life. If the reason why I'm saying
- 25:59this is because if you look back at history that has happened so
- 26:02many times, do you understand where I'm coming from?
- 26:05I understand, but I find it very different.
- 26:08I don't think the changes that we are facing now, especially
- 26:12with AI in terms of speed, you cannot compare it from hunter
- 26:17gatherer to farmers. The industrial revolution, the
- 26:19Internet took years. Hundreds of years.
- 26:23Hundreds of years, this is AI is 3 years, 3-4 years.
- 26:26We have ChatGPT now. Well, AI, the speed, well, the
- 26:29speed is dramatically higher, yes.
- 26:31And the impact every single event that you're highlighting,
- 26:38we are dealing with humans. We are not dealing with
- 26:40something more intelligence than us.
- 26:43Now we have another, technically another entity.
- 26:47You are now handing over the management of your organization
- 26:52to another entity, to an intelligent entity that gets
- 26:56more intelligent every day. So I would think probably,
- 27:00maybe, yes, if you look at it 2000 years from now, this is a
- 27:03good moment. Everybody say, yeah, hopefully
- 27:05we did it. But I I think those events would
- 27:08be categorized really differently.
- 27:11Yeah, OK. How much do you like your life
- 27:14today? A.
- 27:14Lot. OK, what do you enjoy from your
- 27:17life? Do you enjoy the comfort of your
- 27:18life? The fact that you have a bed?
- 27:20A comfortable pillow? Sure.
- 27:22But I always that you can order. I always envy my ancestors as
- 27:25well. So it's.
- 27:27Let just answer my questions for a moment.
- 27:29Do you enjoy the benefit that you can get an Uber ride and
- 27:33just wait for half a second and the car is there to pick you up
- 27:36to take you to your destination? Do you enjoy?
- 27:39But there is a butt? No, there is a butt to the Uber.
- 27:41But then there was no Uber and I had to walk from my school to my
- 27:45home. I didn't know there was, I
- 27:47didn't know there would be an Uber.
- 27:48So I didn't have any problem at that time.
- 27:50I wouldn't say oh fuck I. But I, no, I remember the days
- 27:54where you said, oh fuck, I need to walk.
- 27:56I remember those days. No, you're forgetting those
- 27:59days. I remember the days where you
- 28:00were walking for a really long time and you were tired and
- 28:03exhausted. And the moment Uber came and I
- 28:06seen many nights where you are thankful, you're like, my God,
- 28:09thank God for Uber. The fact that I can just get it
- 28:12in 30 seconds. I got a car you cannot deny.
- 28:14That sure, sure. You are enjoying the benefits
- 28:19that have come at the back of the agricultural revolution, the
- 28:23scientific revolution, the cognitive revolution, and now
- 28:29perhaps after this is going to be the AI revolution.
- 28:31We're not getting the benefits. Well, I actually have to
- 28:34rephrase that. You're already getting the
- 28:36benefits of the AI revolution. I see it on a day-to-day basis.
- 28:40And the reason why we're seeing this benefits even quicker than
- 28:45before is because we're seeing the dangers at the same time.
- 28:48Sure, I actually agree with you on that.
- 28:49This is so much faster that within the first two to three
- 28:53years, you're already seeing the huge benefits and the downsides
- 28:58of the AI revolution. You're right, Previously the
- 29:00agricultural revolution took hundreds of years for you to see
- 29:04the benefits of it and also to see the downsides.
- 29:08But now you see all of it in two to three years.
- 29:10And guess what? The benefits and the downsides
- 29:12have been multiplied by 100. XI agree.
- 29:16Because we now are dealing with a more intelligence, arguably
- 29:20entity. Arguably, definitely nobody
- 29:25agree with that. Now let's discuss the actual
- 29:29impact on the workforce. I completely agree with you on
- 29:33the human and ethical gaps here. The loss of nuance and culture.
- 29:38I have no idea how an AI system is going to be able to handle
- 29:44the human part of the business. Yes, having 10 middle layer
- 29:49management systems can cause lots of bureaucracy, lots of
- 29:55unnecessary back and forth meetings that are uncalled for.
- 30:02That is the good side of the AI. But just like you mentioned,
- 30:07Farhat on his first 30 days was a low level, low tier employee.
- 30:13Low. Level.
- 30:14According to yourself, according to yourself, low score, low
- 30:17scored employee. Had it not been for the trust
- 30:21that your managers put in you, would you have stayed in the
- 30:27company for two 3-4 years to then prove yourself as the
- 30:30exemplary employee that you were?
- 30:33Because this is something I can vouch for.
- 30:36Farah was one of the top in his company.
- 30:38His CEO and the founders were always in contact with him,
- 30:42asking him for advice, and he was managing the biggest
- 30:44projects. If he was up to the AI assistant
- 30:48or the AI intelligence unit, they would have fired you very
- 30:52early on and you probably wouldn't have stayed because you
- 30:55would have had much less friends in the company.
- 30:59How can they replace that? I'm unsure.
- 31:02And the next thing that is going to be even more difficult is the
- 31:06transition, the amount of layoffs, the lack of motivation.
- 31:10I mean, if you see 30,000 people just disappear like that in your
- 31:14company, how can you be motivated to work in that
- 31:18company? The next thing I want to talk
- 31:20about is job security. Guys, the days of job security
- 31:25are long gone. Let me tell you about a story.
- 31:29As immigrants, for all of our lives, one of the main things
- 31:35we've been taught from our grandparents to our parents has
- 31:39always been you want to go to university, you want to get a
- 31:43degree, and then you want to go and work in some of the biggest
- 31:46companies in the world, like Google.
- 31:47If you secure a job at Google, you're good and set for life.
- 31:52That shit is no longer true and you're seeing it in real time.
- 31:56The fact that you can go to Oracle and tomorrow 30,000
- 31:59people are laid off, what happens to that fucking degree?
- 32:03What happens to that lifelong dream that you're going to work
- 32:06in one company and stay there until you get retired?
- 32:10It's gone. That is gone.
- 32:12So I wrote here to myself. I said, if one is afraid, what
- 32:18should one do? Why did I write this?
- 32:21Because I saw a comment by you guys the other day which made me
- 32:25laugh. You said Bad Decisions podcast
- 32:28discussing the world's scariest topics in a funny way.
- 32:31And we try to keep it light and keep it humorous because you're
- 32:36trying to keep it light. We try because some of the
- 32:39topics that we discussed are incredibly exciting and others
- 32:43can be scary. And that is the truth of where
- 32:45we are today, and we're not going to change that.
- 32:48The whole point of this show is not to come and sugarcoat what
- 32:51is happening because that's not going to be valuable for you.
- 32:54And you're giving us your time and we value your time, and it
- 32:57is not valuable for us. Why would we spend an entire day
- 33:01in a week on talking about random things that do not bring
- 33:05anybody value and are untrue? So then I try to think for
- 33:09myself and say if I'm afraid or if you're afraid because you see
- 33:13all of these headlines, people getting laid off. 30,000 here,
- 33:187000 there, 40% of the workforce.
- 33:20What should you do? Well, if you're a freelancer or
- 33:25an entrepreneur, or you work in a small team, not as part of a
- 33:29bigger company, but the company is relatively small, in my
- 33:33opinion, you're actually not in danger.
- 33:36In fact, you have a huge opportunity here.
- 33:40You should look at this article, read it in detail and treat it
- 33:43as an experiment. Don't look at it as a religion,
- 33:47as in you have to follow it because this is going to be
- 33:49exactly how every company is going to operate.
- 33:52No, even within the article, Jack and the other writer, they
- 33:56mentioned that this may face lots and lots of issues.
- 34:00It will require experimentation for them to figure out what
- 34:03works and what doesn't. And so don't just look at the
- 34:06article and say, OK, I'm going to change how I work tomorrow,
- 34:08but look at the article and understand that AI is changing
- 34:12and shaping the way people work and see how you can implement it
- 34:17in your workflow. I wrote this for myself.
- 34:20I said start by mapping your own workflows and pipelines, the
- 34:25team's communication channels and customer signals.
- 34:28So if you're working with clients or customers, try to
- 34:32understand and map out the entire map of the company and
- 34:37then see is there prototypes of AI workflows that can help
- 34:41reduce friction and increase speed.
- 34:45That is essentially what this article is stating.
- 34:47If you remove the layoffs and everything, what the article is
- 34:51stating is can we use AI to increase speed and reduce
- 34:55friction? And if you can do that for your
- 34:57company, if you're working in a small company or if you're a
- 35:00freelancer in your own pipelines, I genuinely believe
- 35:04that this new age of AI can actually help you win more so
- 35:10than if AI wasn't there. And the next thing I wrote to
- 35:14myself and for you guys, it is about developing new skills and
- 35:19upgrading yourself. This is not a time to sit by and
- 35:23let all of this happen in front of you, especially if you're
- 35:26watching this show and not experiment.
- 35:30Every single day of bad decisions is an experiment.
- 35:35Every single day, me, Farrad, and Pratt are throwing ideas
- 35:39around and we implement these ideas almost instantly.
- 35:43If one person has an idea for changing the communication
- 35:47channel, as an example, or building a software that can
- 35:50help upload videos instantly, we test it immediately because we
- 35:54can now because of AI. The worst thing to do is know
- 35:57that these tools exist and not try.
- 36:01The cost is unbearably low. Unbearably low.
- 36:04You should really consider all of the things that we just
- 36:09mentioned and see how this could impact you.
- 36:11And this is regardless of what industry you're in.
- 36:14We are not talking about oh go and try video generation or
- 36:17image generation. If you are in sales business, if
- 36:21you're in marketing, you can automate your process.
- 36:24You can use an AI agent, you can use codecs, you can use cloud
- 36:28code to start getting market analytics.
- 36:32Then start automating a part of your workflow.
- 36:34For example, customers sending you e-mail, you have to store
- 36:37those emails, you have to generate invoices, you have to
- 36:38send them to the customer. Whatever take up the time of
- 36:42your day, you can use AI to improve it.
- 36:44I'm sure there is one tool out there that can help you.
- 36:48We have created 6 apps in the last 2-3 months.
- 36:51I'm sure we can create one and it is important to experiment
- 36:54because when we started cursor lovable was good, then it became
- 36:58codecs, now it's clod. Clod has been stupid for the
- 37:02past one week. I switched to codecs.
- 37:04You have to be experimenting with all these different tools.
- 37:08And for people who work in massive companies, and I'm sure
- 37:13some of you are companies like Epic Games, Oracle, Google or
- 37:19others, I can't name every single company in the world, I
- 37:24would, if I were in your position, really consider
- 37:27upskilling myself and having a backup plan.
- 37:30Simply looking at the market, I'm seeing more and more news
- 37:34like this. And these job cuts are not,
- 37:38like, always about cost cutting. So just because the company's
- 37:42doing well and you're sitting there and you're thinking, oh,
- 37:45my company is profitable, and so I don't.
- 37:48Yeah. So I don't need to worry about
- 37:50this. It is not the case.
- 37:52It's not about cost cutting anymore.
- 37:54It's about the new way of work, which is why we even really
- 37:58wanted to talk about this article.
- 38:00Which means even if you're working in a big company, even
- 38:03if you are in a relatively high position, even if your company
- 38:06is profitable, there are chances that one day they decide that,
- 38:10hey, we can do the exact same job with a smaller team if we
- 38:14add an AI. That is a reality that is
- 38:17happening in the market. And it starts with the big
- 38:19companies like Oracle. But it has a ripple effect.
- 38:22I guarantee you, you're going to see more and more companies
- 38:24follow suit if these fight companies find success.
- 38:28And that's something we'd have to watch for now, I do want to
- 38:32address something. So Pratt has pulled up this
- 38:34ai.bat Decision studio down here.
- 38:37I want to be very honest about what we're doing here.
- 38:39You see, because of this show, we're getting a lot more market
- 38:43sentiment within bad decisions, which means we understand not
- 38:47only what is happening in the world because we talk about it,
- 38:50but we also understand where you guys are From the comments that
- 38:53you give us. We understand who's the
- 38:56demographic of our audience. You guys, we understand some of
- 38:59you are 3D artists, some of you are videographers, some of you
- 39:02are working in big companies, some of you are freelancers,
- 39:05others are students, some are just enthusiasts.
- 39:08You guys are literally just playing around with these tools
- 39:11because you like technology. We're getting more and more
- 39:15insight as to what you guys need and what the market needs.
- 39:19And that's why when we're looking at these news pieces,
- 39:22we've already mentioned this during the last few episodes
- 39:25that we are actually working on developing the platform that is
- 39:30going to teach every single person in this show how to use
- 39:35AI tools practically. And I think this is very
- 39:38important, especially as we discussed this topic today.
- 39:43Why? As you know, the world is
- 39:47changing really fast. There is tons and tons of AI
- 39:50tools out there and the reason why there's tons of AI tools out
- 39:54there is because we now can use AI to help code and so
- 39:59developers who would have taken years to make an application can
- 40:02now only take weeks and make the same application.
- 40:05Non developers like us are making their own applications.
- 40:09So you're going to see a lot more tools flood the market, 90%
- 40:14of which are garbage and useless, but there is the 10%
- 40:18that will be incredibly useful. These are tools such as codecs,
- 40:235.3 clod code. All of these tools have helped
- 40:28us not become incredibly, not only become incredibly faster,
- 40:33but also work with clients in certain domains that we never
- 40:37worked in before. This is confidential
- 40:40information. I cannot share it because it's a
- 40:42current client we're working with, but we're stepping into an
- 40:45industry that is far from anything bad decisions has ever
- 40:49touched. It is completely non tech savvy,
- 40:53but we're marrying it with technology only because AI
- 40:56allows us to do so. And so we're seeing so many
- 40:59opportunities in the space. And for anybody who's been
- 41:02following Bad Decisions for a long time knows that we pride
- 41:05ourselves in education. For the longest time since IT
- 41:09dawn of the Bad Decision channel on YouTube, we've been focusing
- 41:13on education. We use tools, we experiment with
- 41:15them, we learn, and then we teach you guys because we know
- 41:19the landscape is changing really fast and that there's been
- 41:22requests and demand from you guys.
- 41:24We've been working on a platform that will be constantly updated
- 41:31with the latest tools and educating you guys on those
- 41:34tools so that you don't have to spend hundreds of hours like we
- 41:37do on experimentation. You can come and pick out
- 41:40exactly what is right for you. Whether you need to code an
- 41:44application, make a website, make a video commercial, make a
- 41:49photo ad, write an e-mail, whatever it is will teach you
- 41:53all the best AI tools out there. And I want to end it with this
- 41:58one thing. What's amazing is that as we're
- 42:03working on this show, we're actually developing the script
- 42:07for the different categories of this program and we're updating
- 42:10it every single episode. So every single episode where we
- 42:13come here and we talk about these tools and the news, we're
- 42:17thinking, oh shit, this platform can be better if we add in or
- 42:20remove certain tools. And trust me, it is becoming one
- 42:25of the best pieces of work we've ever done.
- 42:27So I want you guys to have early bird access because you're
- 42:31watching and supporting the show, because early birds will
- 42:33come in at a much better deal than anybody else who's not
- 42:36watching the show. And the only thing I'll ask you
- 42:38guys if you're interested is go ahead and leave your name and
- 42:41your e-mail here so that we can put you in the list and contact
- 42:44you the moment we're going to get ready for launch.
- 42:48That was great. I genuinely mean that after
- 42:51looking at everything that is happening, I think this is going
- 42:54to be something the solution. This is going to be the solution
- 42:57because if you are looking to develop new skills and upskill
- 43:01yourself as we discussed, this is going to be our answer to
- 43:04that problem. Because what Paris is actually
- 43:07really right. Companies are in great financial
- 43:11positions to hire, but to hire skilled people, Oracle, Meta,
- 43:16even Epic Games, they are not losing money, they are not
- 43:19cutting costs. Like you mentioned, they're just
- 43:21restructuring. OK, now AI can do this.
- 43:24This person with AI can now outperform 5 employee without
- 43:28AI. So now we want to hire these
- 43:31people. Now in the next topic we are
- 43:33going to talk. One of the biggest AI companies
- 43:35raise billions of dollars. So it is really an important
- 43:40time to upskill yourself. I wish I could find this on this
- 43:44show. If I don't, I'll bring it up on
- 43:46the next episode. But speaking to what he just
- 43:49said, Epic Games did a round of layoffs recently.
- 43:51Yeah, yeah, 1000. Everyone's sentiment around it
- 43:54was, my God, they're losing money, they're cutting Commons.
- 43:57Billions on Fortnite. I just saw a person talk about
- 44:01how they were hired for a significant role at Epic Games.
- 44:04So they're not just laying people off, they're actually
- 44:07hiring as Farad said they are. Restructuring.
- 44:09They restructure the skills, right?
- 44:11I want Pratt, who knows codecs, who can make a website, who can
- 44:15make an app, who can do 3D design, who can generate audio
- 44:18and video and text. Yes, and this rather than like 5
- 44:21different people. This person I forgot, I don't
- 44:23know he who he is because I just found out about.
- 44:26It he was recently hired. Recently hired, just literally
- 44:29yesterday he posted that he is a founder of a company and he's
- 44:33being hired for a very strategic and important role in Epic
- 44:37Games, which is going to be helping them push what I imagine
- 44:41is related to digital characters and AI mocap in Unreal Engine.
- 44:46I don't know exactly, I'm speculating here, but he was a
- 44:49very important role and he is not essentially leaving the
- 44:53company that he's. He's like a consultant.
- 44:55Yes, exactly, which is why I'm saying these companies are not
- 44:59simply getting rid of people for the sake of cutting costs, they
- 45:03are restructuring to people who can help them work in this new
- 45:07age of AI revolution, but that's about it.
- 45:11I would love to hear your opinions on what you think about
- 45:15the article where you think this is all going.
- 45:17Please let us know in the comments.
- 45:19I read with Farrad and Pat, every single one of you, whether
- 45:23you're on YouTube X or Spotify, so your comments matter a lot.
- 45:28Thank you. Farrad, are we ready for the
- 45:30next topic? Let's Go Open AI raises $122
- 45:35billion to Make the AI Super app.
- 45:40You see, as models become more capable, the limiting factor
- 45:45shifts from intelligent to usability.
- 45:48At the end of the day, Claw, Gemini, Grog, Open AI, they all
- 45:53have intelligence model, but the company wins who can get user in
- 45:58one single platform and give them everything.
- 46:01Because users don't want this connected app.
- 46:04They want to go into one place that understands them, that can
- 46:08take action, that is connected to all of their workflows and
- 46:11datas and apps so they don't have to go outside of it.
- 46:15That is exactly what Elon Musk is trying to do with X, to make
- 46:19it the app of everything, a payment platform, a messaging
- 46:22platform, a social media. Of course, an AI model called
- 46:25Grok Open AI have the same vision.
- 46:29And that is why they raised $122 billion.
- 46:32They want to bring ChatGPT codecs, their browsing features,
- 46:37and all their agentic capabilities into one location.
- 46:42When they did their $122 billion raise, their valuation was $852
- 46:50billion. To put in perspective, this
- 46:54valuation is bigger than the annual GDP of Belgium, bigger
- 47:00than the annual GDP of Singapore and bigger than the annual GDP
- 47:05of Ireland. You just said GPT of Ireland.
- 47:09But a company which is not very old.
- 47:14Not that old. The valuation is bigger than a
- 47:17country's annual GDP. That is nuts.
- 47:20They raise this money to, of course, fuel AI infrastructure,
- 47:24compute expansion and do next phase development.
- 47:27It was raised by Amazon, NVIDIA, SoftBank and Microsoft and of
- 47:32course A16Z. Some might ask that person is
- 47:38me. I asked you the first time when
- 47:39I saw this news in the morning, why would they pay them $122
- 47:43billion? Who evaluated them at 852 what
- 47:46they have done? Sam Altman says Open AI,
- 47:51exclusively ChatGPT is the fastest AI platform to reach 10
- 47:56million, the fastest AI platform to reach 100 million, and soon
- 48:02the fastest AI platform to reach 1 billion active users, 1
- 48:07billion active users weekly. Their revenue when they start in
- 48:11the first year was $1 billion from ChatGPT.
- 48:15It became $1 billion per quarter, now 2 billion per
- 48:20month. So they made 1 billion to 4.
- 48:23Billion a year and then 2 billion per month.
- 48:26Per month now it's 24 billion per year.
- 48:30Fuck. So they're making shit loads of
- 48:32money. And ChatGPT itself has 900
- 48:36million weekly active users. That's close to a billion.
- 48:40That's 1/8 of the earth population.
- 48:43So they not only they have revenue, they have a large
- 48:48number of active users. Their total users for their app
- 48:54is bigger than the other four app combined.
- 48:57So you can see. Yes, the next the total AI time
- 49:02spent in ChatGPT is bigger than the next 4 app combined, so
- 49:08maybe those. Next 4 apps are.
- 49:10Probably Clod and Grog and all those.
- 49:12So Clod and are great at enterprise, but when it comes to
- 49:15mass market, ChatGPT is huge, a great stat.
- 49:19The ads pilot that they run on the free users.
- 49:22So you see a lot of free users like our parents are on ChatGPT.
- 49:26The ad revenue brought more than $100 million annualized in just
- 49:31under six weeks. So it shows that OK, we gave a
- 49:34lot of people for free, they're using our compute let's run ad
- 49:37for those people and then we charge the premium members.
- 49:40So now both groups will give us revenue.
- 49:42So it's they made it very simple, right?
- 49:47We need more money. We need more money for more
- 49:49compute. More compute will give us more
- 49:52intelligence. More intelligence will give us
- 49:54better products. Better products give us faster
- 49:57adaptation, of course, more revenue and more cash flow.
- 50:00So you give us your money, we make back your money, as simple
- 50:04as that. But another important factor, I
- 50:07think that I was reading the statement that Open AI wrote
- 50:11about this round of money that they were raising, a lot of
- 50:16money would be spent on compute, on the infrastructure.
- 50:20So they are working with NVIDIA, of course, that's their main
- 50:23chip partner, but they're working with Microsoft Azure,
- 50:25they're working with AMD. All this money will go into this
- 50:30AI companies, at certain point, they have to reach a revenue
- 50:34level to pay it back. When do we reach that level?
- 50:38Are we going to get more expensive subscription prices?
- 50:43Are we going to get more products that we need to pay
- 50:45for? Maybe at Open AI physical
- 50:48hardware app? We have to see.
- 50:50But I think the first, next phase for Open AI is to make the
- 50:55everything app, no more ChatGPT, codecs, Atlas, the browser,
- 50:59everything in one place. What do you think about?
- 51:02This. First of all, that's a massive
- 51:05valuation. Have you seen a larger
- 51:08valuation? What's the largest valuation of
- 51:12a company ever? I'm curious what that number is,
- 51:17because $122 billion. No, but you cannot say because
- 51:22of course Tesla is a trillion dollar company, right?
- 51:25They combined right with who? No, that was XA and SpaceX.
- 51:30No, I think it's NVIDIA probably #1 number one company on stock
- 51:34market should be the one. I think it's either Apple or
- 51:36NVIDIA, Probably trillion. $122 billion, but you're right, the
- 51:41fastest company to get there. They're the fastest company and
- 51:44bro, they are four years old and they are a software company.
- 51:48So yes, NVIDIA is 5.37 trillion, but it's a public companies.
- 51:52These guys are not even public. What do I think about it?
- 51:59I'm not surprised, yeah. I'm actually surprised at the
- 52:02money that they raised, but. But I'm not surprised because I
- 52:05know the growth of these applications is unlike anything
- 52:10else. Yes, I read that the growth of
- 52:14these applications compared to Facebook and all of these
- 52:18different applications is substantially faster and the
- 52:22fact that they need all of the money as well to grow and
- 52:27compete, it makes perfect sense. One, compute is expensive.
- 52:32We all know that. We know that these AI models,
- 52:36the inference and the usage of them will require lots and lots
- 52:40of power and resources. In order to create smarter
- 52:45models, they need to train them. Training models is incredibly
- 52:51expensive. Using the models is also
- 52:55expensive. So you have the training when
- 52:58you build the next GPT 5.5 or 6 point O, that will require
- 53:03compute, which is being done invisibly.
- 53:06This is being done without us even noticing.
- 53:09Then there's a part that we notice is the part that we use
- 53:11Chachi BT when your mom is asking for a cooking formula and
- 53:15recipe or when you're asking it to draft you an e-mail.
- 53:19All of that is using compute. It's a cost for open AI and just
- 53:22like you mentioned, there's a huge user base that is living on
- 53:26the free membership. Which now they're making money
- 53:29off ads from that. Using ads and those ads are only
- 53:31going to increase, just like we've seen with YouTube.
- 53:35Like in the beginning, you're not getting it on the paid
- 53:37version. No, but same like YouTube,
- 53:40right? YouTube you have the premium
- 53:41subscription that doesn't have ad, but if you are on the free
- 53:44membership free tier, you're getting ads back-to-back.
- 53:49A pre roll, a mid roll, multiple mid rolls and you get ads on the
- 53:54sides. I mean, I don't have any
- 53:57specific feeling towards this. I know that this space is going
- 54:02to be getting the largest investment ever in terms of
- 54:05where we are today. The AI space is receiving the
- 54:09largest valuations. And the investments, with this
- 54:12investment, they can beat Claude Anthropic.
- 54:14I'm telling you, chatability codecs last one week much better
- 54:19than cloud. Guys, I'm not kidding.
- 54:21We are all using both. We are on Cloud Code Max 20X,
- 54:26ChatGPT Pro, whatever it is. All of our apps are connected to
- 54:30both cloud and codecs. From beginning of March until
- 54:34last week, all of us combined Cloud code, hands down the best
- 54:39last week till now. Codecs, I don't know if you have
- 54:42experienced that, I haven't. Codecs really good.
- 54:44I showed you the UI that it created yesterday.
- 54:48Let's call it a close race, no one is the ultimate winner.
- 54:53However, Claude has been leading for the past few weeks,
- 54:57especially after the Department of Defense conflict where many
- 55:01users switch to Claude. Claude's user interface is more
- 55:07premium in my opinion at the moment, especially in Claude
- 55:11chat. I don't know if you guys have
- 55:12tried it recently. I asked it to design me a brand
- 55:18and within the chat it is whipping up sliders and buttons,
- 55:23dynamic elements, graphs. Instead of reporting to me with
- 55:28text, it's making graphs live in real time.
- 55:31So it's coding graphs, visual interfaces, just to respond to
- 55:35me. Claude is the only one doing
- 55:37that, and I feel like the market sentiment is shifting slightly
- 55:41towards Claude, especially on Instagram where there's a lot of
- 55:45creative and creators. More people are using Claude
- 55:48now. So the user base is shifting,
- 55:51but I don't think it's a massive win for Claude.
- 55:54We've seen something incredibly crazy in the AI space, which is
- 55:58the moment you think a model, whether it's image generation,
- 56:02LLMS, or video generation, you may think they're winning.
- 56:06Next week another model is released that completely takes
- 56:09over their user base. I agree.
- 56:10So I wouldn't marry myself to any single tool just yet.
- 56:15You know, there was a time where you could.
- 56:17Confidentially married to a bad decision, yes, that will give
- 56:20you that's. Not a tool.
- 56:21There was a day where you could confidently say I'm done using
- 56:25Yahoo, I'm going to go to Google, and that day actually
- 56:28occurred. That day happened.
- 56:30But it's too early to do that with AI.
- 56:31We may never even have that with AI.
- 56:33We may just continuously get a flow of apps and softwares and
- 56:37models and have to switch and adapt.
- 56:41I don't know. But one thing is for sure, I
- 56:44don't think throwing money alone at compute is going to make
- 56:48ChatGPT win. To answer your question.
- 56:50Throw money at me, I'll be better.
- 56:53Throw $122 billion at me, I'll be the best.
- 56:56But. Claude also has money, true.
- 56:58It's not like. Also recently how much anthropic
- 57:01raise it's. Not as if Claude doesn't have
- 57:03money and Elon is saying that the next versions of Grok code
- 57:09and grok imagine for image and video generation are going to be
- 57:12epic. Actually, anthropic.
- 57:14So 30 billion series by the value, yeah, by the valuation of
- 57:18380 billion. Yeah, half of this, but they
- 57:21race four times. But you have to understand it is
- 57:25the code structure and the training method and the PR and
- 57:31so on and so forth that also matters.
- 57:32Of course, you cannot quantify that.
- 57:35I know, I know. Money purely.
- 57:38You cannot say I'll throw money at my PR team and they're going
- 57:40to do a good job. If your PR team fucks up, then
- 57:43everyone goes to Claude. Yep.
- 57:45Do you understand what I'm saying?
- 57:46Even if they have a billion dollars to spend.
- 57:48Yeah, yeah. Makes sense?
- 57:50For me, this is great news for open AI and ChatGPT users.
- 57:55They're gonna see smarter versions, but if they're gonna
- 57:57win the race, I don't know yet. All right, Farah, shall we move
- 58:01on to the next topic? Let us know in the comments.
- 58:03Are you a GPT user or a cloud user?
- 58:06Yes. Which ones do you use and
- 58:09specify? ChatGPT versus cloud chat,
- 58:14codecs versus cloud code if you are coding using any app because
- 58:18the chat and the code versions are very different.
- 58:21So let us know you are Team Open AI or Team Anthropic, shall we?
- 58:26Let's go. Here's an AI tool that seems to
- 58:29be helping 3D artists and game developers.
- 58:32It's called Tripo 3D. Many of you guys might have
- 58:36heard of it. We've perhaps talked about this
- 58:39in the past. It's not the first time that
- 58:41we're covering this, but the reason why we're showing you
- 58:44this video specifically by Royal Skies, which is one of my
- 58:48favorite Youtubers on Blender tutorials and Unreal Engine
- 58:53tutorials. I wanted to show you this video
- 58:55because he's covering Typo 3D and there's a specific thing
- 58:59that he's showing which gets me as a 3D artist, excited.
- 59:04So let's watch this for 30 seconds. 3D artist on the planet
- 59:07has been asking an AI company to try and tackle the issue of
- 59:09retopology since the dawn of time and today I am happy to say
- 59:13that we finally have a sponsor that has decided to try their
- 59:17best to tackle the issue for us once and for all.
- 59:20Four months ago, Tripo asked my community to stress test their
- 59:23software and provide feedback. So we live streamed an event
- 59:27where me and 27 of the best 3D artists and game developers in
- 59:30our community got together and we basically tested the hell out
- 59:34of their software and we actually broke.
- 59:36It and it was. A lot of fun and we gave them a
- 59:38big log. Yeah, you can stop the video.
- 59:42Royal Skies has been making videos on YouTube for a very
- 59:44long time. He makes incredible tutorials
- 59:47for Blender, for Unreal Engine, for game devs specifically.
- 59:52We did mostly for cinematics, but he does it mostly for game
- 59:55devs. But the reason why I really
- 59:57enjoyed this video although it is a sponsored video by Tripo.
- 1:00:01He did mention a couple of things that stood out to me.
- 1:00:04Tripo worked with him and 27 other game devs and 3D artists
- 1:00:11to help stress test their tool, and this is the most significant
- 1:00:16thing for me. A lot of these new AI tools that
- 1:00:18are popping out for creatives are simply tone deaf.
- 1:00:22They're not working because the guys behind the programs are
- 1:00:26simply developers trying to make money.
- 1:00:28They're not creatives at heart. And if and if you're not a
- 1:00:32creative at heart and you're a developer and want to help 3D
- 1:00:35artists and game developers, you have to take the triple 3D
- 1:00:38route. The fact that they handpicked
- 1:00:42and selected Youtubers, creators, 3D artists, and game
- 1:00:45developers to come and stress test their tool and give them
- 1:00:49feedback and that they were open to feedback is the most
- 1:00:53important thing. This is how these companies and
- 1:00:56tools can actually develop functions that help us not
- 1:01:00replace artists, but help artists.
- 1:01:03And this is one of those great examples.
- 1:01:05So what he's actually showing in this video is his early tests
- 1:01:09with Tribo 3D and the fact that they live streamed with other
- 1:01:12artists, stress testing the tools they send the feedback and
- 1:01:15then they flew out him and other artists to GDC where they were
- 1:01:20reviewing and showing the new version of Tripod 3D.
- 1:01:24And what he's claiming in this video is that there has been
- 1:01:27great improvements made in not just the 3D asset generation of
- 1:01:31Tripod 3D when you give it an image and it turns it into 3D,
- 1:01:35but re topologizing. For anyone who has been a game
- 1:01:39dev or has used 3D models on the web for augmented reality or for
- 1:01:44VR, we know that optimization is a huge part of the workflow.
- 1:01:48And partly that part of the workflow is a huge pain in the
- 1:01:52ass because you want to focus on the creative aspect of making
- 1:01:56the models or making the story. Wherever you lie in that
- 1:01:59spectrum, the optimization is the least creative bit.
- 1:02:03I know some people do enjoy retopologizing.
- 1:02:07I'm not denying that, but what I'm saying is that a lot of
- 1:02:10artists have to spend long hours focusing on Ritopo when they
- 1:02:15really wished there was an automated tool that did it for
- 1:02:19them, and I genuinely believe that AI can help figure this
- 1:02:24out. This is what the video is
- 1:02:25showing. You can go ahead and check out
- 1:02:27the video yourself, it's on X and on YouTube.
- 1:02:30But what I'm seeing is actually quite promising.
- 1:02:32The results that you get out of the 3D models and the retopo of
- 1:02:37the 3D models is pretty insane. Sure it's not 100% perfect yet
- 1:02:42and no human was. I guess at the same time, if you
- 1:02:45compare it to a human who was optimizing their assets, we can
- 1:02:48never claim that a human did 100% accurate retopo of a model.
- 1:02:54Some cases exist, but not always.
- 1:02:57And so just looking at these tools and the developers talking
- 1:03:01directly to artists and creatives developing tools
- 1:03:04specifically made for them, solving their problems is where
- 1:03:08I stand. And although I haven't used
- 1:03:11Tribo 3D as much, I love seeing posts like this.
- 1:03:16Epic Games, Blender Maxon should acquire tools like this.
- 1:03:22Bring it inside the software. Exactly what Sam Ultimate tried
- 1:03:26to solve for Open AI. It's not about the feature, it's
- 1:03:31about usability. If it's within Blender,
- 1:03:34everybody will use it. Andrew Price will put it in his
- 1:03:38new Donut tutorial. Everybody will know how to use
- 1:03:41it, right? So all this big company, I think
- 1:03:43Epic Games did it with mesh Capade.
- 1:03:45I think they required mesh Capade, right?
- 1:03:47All this feature based AI tools should join a big 3D software
- 1:03:54because that's where people live unless they want to make their
- 1:03:57own, which is a different thing. Or what we see Autodesk is doing
- 1:04:01with Wonder Dynamics. Bring different tools, AI
- 1:04:04generation tools within that. I think it would help artists a
- 1:04:07lot if we have everything in one place and they can have their
- 1:04:11workflow much more efficient. I believe that is going to
- 1:04:15happen. My only question is when how
- 1:04:18fast are these 3D softwares going to implement tools like
- 1:04:24this? Let's not forget ZBrush and
- 1:04:27other tools already have retopo tools.
- 1:04:30It's not like they don't have them, it's just sometimes
- 1:04:34they're incredibly complex to use, especially for a beginner.
- 1:04:37But imagine if you're just starting out on your 3D journey
- 1:04:40and you get it and then there is an AI button that helps you
- 1:04:43optimize your asset immediately. Then you.
- 1:04:46Can all this 3D software should have a light version, right?
- 1:04:49What is a light version like a Higgs?
- 1:04:51Field for a 3D artist. Something that I can come in
- 1:04:55with 03D knowledge. I can drop an asset, put on a
- 1:04:59camera, put on a light text from my animation, text from my
- 1:05:03environment. Yes, I mean Unity is doing that
- 1:05:06right. Unity has the Unity Studio on.
- 1:05:08The web. I think that is the way if you.
- 1:05:10But Roblox is also doing Unreal also has it for architecture.
- 1:05:14The twin motion. Yes.
- 1:05:17No, but you have to understand these tools are designed for
- 1:05:21advanced users. They don't.
- 1:05:23They don't intend to. Serve you.
- 1:05:26Who wants to just come in and drop in a light?
- 1:05:29There are web apps that do that already, sure.
- 1:05:32So I don't think that's necessarily the solution.
- 1:05:35I do believe that they should continue serving the advanced
- 1:05:38users by providing them with tools that help them work
- 1:05:41faster. If you enjoy doing Retopo, I'm
- 1:05:44not asking you to go and click a button and have it done, because
- 1:05:47that's probably not going to be fun for you.
- 1:05:49But I personally know a lot of game developers who can focus on
- 1:05:54creating better games if they can make it in a faster time
- 1:05:58because that means they can spend less money and therefore
- 1:06:01reallocate that money to areas where it matters for them.
- 1:06:06This is a promising technology in my opinion.
- 1:06:08Let me know what you guys think in the comments.
- 1:06:09I know for a fact a lot of people are actually testing
- 1:06:14these tools, especially you guys watching.
- 1:06:16So please let me know because we don't actually test all of them.
- 1:06:20There's way too many. But if you tell us what you
- 1:06:22think is really good, we'll put it in our work flows and then
- 1:06:26we'll bring it on the show. There's actually one guy at
- 1:06:29Pratt. Can you pull up Halim's post on
- 1:06:31Instagram? Yes.
- 1:06:33One of our students from the Ultimate Unreal program is
- 1:06:37continuously making new animations, and he's actually
- 1:06:40implementing some of these AI tools.
- 1:06:42He's an animator by heart, and now he's in Unreal Engine using
- 1:06:46these new AI tools to help him create characters.
- 1:06:49And this video that you're looking at, so the characters
- 1:06:52that you're seeing in front of you there, he actually used AI
- 1:06:55to help him with the ideation and then created every single
- 1:06:59part of the character, the belts, the gloves, the head, as
- 1:07:02you can see with different prompts and then turned those
- 1:07:05into 3D assets. Of course, he went ahead and
- 1:07:09refined them and optimized them himself manually, perhaps even
- 1:07:13edit them further. I'm not sure of the entire
- 1:07:15workflow, but I watch a couple of these videos.
- 1:07:17You can see Tripo there as well. Yes, use Nano Banana for the
- 1:07:20images and Tripo for the 3D. And if you show the final video,
- 1:07:24Pratt, I really like the final video he made.
- 1:07:26The final video with the animation manually made inside
- 1:07:30Unreal Engine. Look at this video.
- 1:07:32So cute. It's insane.
- 1:07:37So he used AI to speed up his workflow and make this faster.
- 1:07:41The parts where he loves doing and wanted to manually have
- 1:07:44control like this, animation, keyframes, he did it himself.
- 1:07:48But some parts he used AI. This is the way I think AI and
- 1:07:523D are going to work hand in hand to help artists.
- 1:07:56Pretty cool. Check out Halim by the way.
- 1:07:58He's an amazing artist and I believe that brings us to the
- 1:08:03end off today's show. I just want to say thank you
- 1:08:06everybody for watching and listening to this episode to
- 1:08:09episode 101. Thank you to all of you and we
- 1:08:14will see you guys on Friday. All right, TGIF.
- 1:08:18Ciao.