That whistleblower didn’t expose anything we didn’t already know though. It’s not a secret that they use copyrighted data to train their models. His death is suspicious, but I’m not convinced yet that it was a hit by OpenAI. I do hope they investigate it more
He probably contributed to copyright infringement while working there or saw negative effects in family and friends of what he was helping build. Another issue is that engagement algos can be very reflective of what someone is saying, so an accuser can get the feeling that the internet is accusing him back.
@2beJT Yeah, no. Not everything is so cut and dry as what you might initially think. It's pretty easy to call any type of takeout as self takeout. These people do not play by the same rules.
bc they can't keep their mouth shut.. Karen Silkwood was the first WB if I'm not mistaken, there are several podcasts you can listen to about her story its not a coincidence for sure
The worst part is that this happens so often nowadays that you can never possibly tell if it was an accident or not. The fact that it happens so often... makes you scratch your head.
Something to consider that seems to have been omitted about the results shown at 8:35, is that the OpenAI model was allowed 50 submissions per question, while each of the human participants were only allowed a single final submission per question. This was also the case in the Math trials as well, where the OpenAI models were allowed hundreds or even thousands of attempts to solve the given problems. So while it is still very impressive that these machines are able to solve these problems in a comparatively short timeline when compared to a human, these are still well-structured problems with distinctly defined solutions. While "in the wild", and especially with non-trivial professional use-cases, AI output frequently requires heavy oversight from a field expert to analyze and verify its validity. Anyone whose toyed around with text, image, song, and code generation knows that it can require multiple attempts to get the desired output (or in some cases, even anywhere in the ballpark of desired output), and the more complex and technical the task, the higher the level of required expertise and scrutiny is needed to make sure the output is correct (or can be modified to be correct).
You talk about Humanity like you don't have any blame... do you have an iPhone, Apple product, or anything made in a foreign country by slave labor?... think about it... "Humanity" is a "concept", not a REALITY... there are still actual Slave Markets in parts of Africa and the Middle East... there are illegal "slave style" workshops in most Western Countries using illegal migrants and children... Ideology and REALITY are two totally different concepts...
@@thecrackin-u8pI don’t even think making money is Altman’s goal. He has become one of those movie supervillains who just wants to see how further he can push this.
@ScaryYokai but rest assured very very bad people will be using true ai, (kinda are now but ai isnt their yet, talking full Ai, when its teaching itself) not this artsy shit....in 20 years it's gonna be interesting times.....
are they "stealing" content from the internet? Yes. That's the answer... Yes. Let's stop deluding ourselves as if original data is being created. It is not.
It's more a question of what loophole in copyright and/or ToS they found, with a few backroom deals between themselves and websites probably being kept quiet.
Well... synthetic "inspired" data is being tested out as a supplement to training data or as a full replacement, but the research won't lead to anything since synthetic data needs human curation to be usable, which defeats its whole purpose. So the argument will hold for as long as generative AI exists
They aren't. The AI uses data from anything the user provides regardless of their copyright. The Indian guy committed suicide because his stupidity got called out. ChatGPT is a tool. Whether a user is responsible is none of their business because that is impossible to control.
Imagine being a worker at OpenAI and having even a 1% doubt in your head that the company you're working for might have killed an ex-colleague? Not exactly a great work environment 😮
OpenAI wants to be regulated by the government so the regulations stop/crush it's competition. Lots of ai startups right now - will be interesting to see how it plays out.
Exactly. These corporations have met the threshold that any regulation on them only hurts us exactly like how copywrite laws have done. Once you amass so much information and wealth on something it's hard to take that away from them even with regulations. Disney still has the back catalogs of like 80 years of drawings and countless of scripts.
Maybe because there is a long track history of people believing and even those who don't believe eventually forgetting and moving on. Unless you are rich or famous most people don't care.
Been rooting for this turnaround in opinion ngl, I mean these guys literally want the eradication of art as human expression its hard to get more evil and soulless than that.
Neoliberalism ruins it every time. Pick a single problem in American business/politics/economics/etc. profit before people guarantees the worst outcome
@@NewMitchell-wh3fj You don't understand AI word smoothy? Just take a buzzword, throw in some hyperbole, some words from a glossary, blend, and top with some more hyperbole. What it lacks in nutrition it makes up for with ... stuff.
No. Actually it means that considerable pay-offs, bribes, and political influence has been so integrated into our financial system that many businesses will fail as a result of one failue.
AI scoring high in tests is because it have access to all the necessary repository... is the same as a person doing these exams with a laptop and internet access to search for things...
Yeah, AI evangelists pointing to these tests when they're like handpicked tests that the AI particularly excel at. The problem is almost none of these actually reflect real world usage
We won't reach AGI until truly logical models surpass us. Generative AI can only be as good as the data it's given, which can't be higher quality than the data us humans have produced
@@rishitkhanna336the information needed to come to a conclusion is still available though. there may not be outright solutions but they still have a lot of info to work with.
Ai corporations are directly profiting from everyone’s work without even paying them anything or asking creators all while convincing the layman to think ‘oh well it’s the future bruh adapt or die.’ Nah, it’s pretty much theft.
I have been watching ur channel for about 3 years, your tsmc/invida video changed my life financially. I cannot thank you enough dagogo! Keep going cuz!
@bryantony9836 well it was back before invidia's most recent surge . The video goes into invidias partnership with semicondu tor manufacturer in Taiwan TSMC .
@saadderaiya both . Tsmc was flat already and I was already in invidia , but that video convinced me to close some other positions and get heavy into invidia . I'm not super investor savvy or nothing but in. Idia already was showing impressive growth pre and post pandemic so it made alot of sense when tsmc started using invidia's tech along with the AI boom. Sorry for long response.
OpenAI went full Boeing quite fast. Though main problem is, that all these companies are breed on awesome, capital intensive and hype prone US market - in tthe same country, that has Congressional Hearings for Big Tech but also on Aliens / UFOs.
Yea those hearings about not UFOs, that word is dead, but like Anonamlyous Flying Entity AFE? will be interesting, specially while Trump is in office, he could reveal long hold secrets (which sure exist but idk to what extend) 👀
From the firing of and rehiring of their current CEO Sam Altman to the death of death of former employee and whistleblower Suchir Balaji... THE WORSE IS YET TO COME
A note with o3. While it wasn't trained on the ARC dataset, it isn't default o3 either, but a version fine tuned on the ARC public examples. They also spent significant compute resources estimate to cost $1-10k per task.
Feels like the old days during the browser wars. One company creates a new feature - the rest copy it and/or improve upon it. Pretty soon the market is reduced to just 2 players - just like everything else in the world.
OpenAI’s most recent products have absolutely not received a Lukewarm yawn from its audience. O1 has been completely amazing. Search, Voice Mode, and other releases made during the twelve days of Christmas campaign were impressive too.
I think the whole legal stuff regarding copyright and training of models is going to be a massive slam on the brakes of all 'above ground' models. Big platforms will probably exclusively license their content for AI training and it will probably cause publicly available models to stop being trained until most of the internet-pie has been sliced. Or they have to figure out a way to mark a source of the trained data so that they can enable/disable it based on shifting licensing agreements. But I'd guess platforms and web application firewalls will soon block or log (for litigation) AI training requests from parties not currently licensed.
You work for 40yrs to have $1M in your retirement, meanwhile some people are putting just $10K into trading from just few months ago and now they are multimillionaires
First, videos are great as always. A small detail at 8:23 : o3 was only revealed as you previously said, not released. Also and on a side note, love the music and “dimmed radio voice” at the beginning and ending, it both feels dream-like and nostalgic.
I love the way ColdFusion explains the things. The most important, the information is unbiased which is very difficult to find elsewhere. Keep up the work! Wish a lot of growth for ColdFusion in the year 2025! 😄💫
the dude is in all ways right these AI uses copyrighted materials the problem is proving it does since its meshed with other stolen materials from other stolen materials its hard to distinguish
This is kinda like when the CIA does some horrendous operation and then the operation is declassified 20 years later. The public goes 'wtf!?' the CIA is like 'what u gonna do bout it cuh?'
There should have been a greater focus on the O series of models in the video. It was the defining breakthrough of 2025 and a much bigger deal than any of the drama. Many who report on AI seem to focus more on the drama than the technology itself. To keep up with the most recent developments in the technology itself, and their broader implications explained in simple terms, I would highly recommend the channel, Ai Explained.
0:50 Executives are paid very highly and would rarely resign such a high paying position. So it's a red flag that the fact top executives(so many of them) are resigning. It means they don't feel safe in their positions, most likely as they are seeing very illegal and unethical practoces that isn't even worth the risk of being sued upon. The fact these people are resigning willingly means there's something sinister going on in OpenAi.
I wouldn't include ChatGPT in the "AI fatigue" category. Image and video production, sure... maybe... (a few communities that I belong to would beg to differ) , but CharGPT has manged to give answers quickly to questions I have had for years that vigorous web searches had not provided.
hello non american 🇺🇸 open ai altman vs musk is a movie 🍿 script writers dream. *PLUS* that person gets to use chatgpt, so that writer will have no problem, just gotta watch 👀 to let it unfold.
I just saw an article that said OpenAI's latest model costs $1000 PER QUERY . That's how much computational power they're using. It's unimaginably unsustainable. The ONLY reason openAI hasn't collapsed is because they have a $65 Billion / Quarter cash pile coming from Microsoft.
@aaronalquiza9680 I also use deepseek/deepthink, it's really impressive, but nowhere close to o1/o1pro. Ofc I must consider the fact that deepseek is practically free or very low in cost. With o3 mini coming in Feb/March and the full o3 coming soon after, the gap is very large
AI is Sooooo good. its probably why Cursor decide out of nowhere to delete entire features from an existing working software. When asking why it can't answer anything logical. We now need to get trained on the pitfalls of AI.
We are in the age of running LLM at home, there are already several LLM brands that are not too user friendly to install, once this aspect is cleaned up, this will take off like a rocket.
@@sirsamiboi we can run LLM like ChatGPT in our home. LLAMA can compete with it but it requires at least 64 GB RAM and a strong GPU last time I checked. There are also smaller models but they are not satisfying as a chatbot.
Ollama2 installs in minutes, requires no accounts to use, runs on Windows, Linux, or Mac. Everything sets up effortlessly, The most user-unfriendly thing is you have to run it in a command prompt. It's very powerful and runs on your own local hardware. I've run it on a thinkpad laptop with no GPU and it will answer queries, sometimes it will take a while to get replies (1 to 2 minutes). On a machine with a modern GPU, queries return instantly always. Absolutely no need to step down to using something like OpenAI's products, etc. and support companies I despise, when I can run it locally, it doesn't even need an internet connection to function and get you answers.
Tax the AI companies now. Massively. That way as a government you'll have more money available to give benefits to those that have lost their jobs to AI in the next 1-3 years and will be claiming under joblessness. It's guaranteed the majority of people in offices right now will lose their jobs to software devs and IT guys maintaining the AI for the companies. They're aiming AI at the housing market which has disaster written all over it. EVERY SINGLE instance of money-saving techniques in history introduced to an industry have been taken on and applied. EVERY single one, every time. Quicker techniques, cheaper processes, whatever it is. "AI can save you money by cutting X% of your workforce, and you get the workforce to teach it their jobs before they leave".
I'm still waiting for AI to develop internal code standards to be transparent and identify itself. Meaning, AI should be capable of interrogating another product (audio, video, text, image, etc) and confirm if it is original or if it is AI generated. This TRUTH should be encoded in the product geocode.
Any model that can reliably detect generated content can be used to train another model that would fool the detector, it's called a Generative Adversarial Network. A perfect "AI detector" can always be used to fool itself
Quite often the originals don't usually win. It is usually the ones who sit back and pick the scraps that win. Opening doors for the one who will sit in the car...
I'm a bit conflicted about copyright law being applied in the same way when used for training data. Maybe they should pay something for using it to train models, but I don't think it should be classed the same licensing content on a platform like Netflix for example, as you're not distributing the content itself, it's more equivalent to how human's learn from various sources, if a person learns something from an article online and goes on to share that information is that breaking copyright law? I feel copyright law would need updating to be suitable for the changes in modern times.
The artists don't wanna accept that though, I think before any progress can be made people need educating.. people need a reality check on where they draw influence from.. I think people want to cling onto the fact that we are more special than a computer program. Every argument you see from artists against this use really undefinable language like "soul" "creativity" etc etc. This ai tech is heading down the path of how our brains work.. it's miles of yet but the principles are already there.. The models are just super primitive at the moment.
I understand the conflict man.These things are very complex. But it's not equivalent as to how humans learn, and this is a dangerous idea. Our leaning process has evolved from millions of years of "information processing", it involves many parts of a body. It's not just the brain. We are not regurgitating statistics like the machines. And we develop style. It's incredible that these LLM's with all the art they consumed can't develop and original, never seen before style. LLM's don't even know what they are doing... And, dude, just why we keep using the human analogy? We should have more rights than machines, more rights that corporations. They are stealing. Let's call it like it is: A theft. They are stealing from one of the most vulnerable people of our societies and by doing so they are depriving them of their jobs too. They employ slave labour as well. I have no words for how shitty this thing is and this is just the beginning! At least Google and many other tech companies took some time before they were the evil ones. These guys are bad form the start.
So I don’t wanna burst your bubble here or anything but my great uncle worked with arpa before it became darpa and he said they were working on LLm’s back in the 1970’s and that our thaad and aegis defensive missile systems were being incorporated back before they were known systems. He said imagine the most advanced thing possible for technology and then go 40 years into the future and that’s what darpa is working on currently. Those “UFO’s” you hear about on the news etc, that’s darpa physically testing systems. Haarp systems pushes extra energy to the crafts when needed for high maneuvering amongst many other things… The world is gonna be just fine, we’ve already given it over to the AI masters, it just hasn’t been allowed to be known officially yet…
I have heard similar statements from people who used to work on black projects. I wonder if there are certain problems with these systems that make widespread adoption not feasible?
This just makes me more bullish on them. When you want to know who's the best you simply need to ask who's getting attacked by everyone from almost all sides for complete bs reasons?
You mentioned "breathing microplastics"... do you know that you are breathing microplastics every time you enter a hot vehicle in the summertime?... that "film" on the inside of your windshield is outgassing plastics from everything plastic inside your vehicle... especially the dashboard... a majority of the dust in a carpeted room is probably microplastics... Those "reusable/refillable" home soda machine bottles have, in small print, an "implied WARNING" by listing and "EXPIRATION DATE" on each plastic bottle... WHY?... Microplastic Deterioration... IT'S ALL AROUND US...
The sacrifices whistleblowers make to expose inhumane corporations..
We would like it to be otherwise but - a human life has a price. And investors don’t tolerate failure.
That whistleblower didn’t expose anything we didn’t already know though. It’s not a secret that they use copyrighted data to train their models. His death is suspicious, but I’m not convinced yet that it was a hit by OpenAI. I do hope they investigate it more
In other words people will do inhumane things for greed.
Or another entity concurring and being angry at OpenAI's leading success and thus knowing who to make die when hearing about a Whisteblower
@@UkSapyy I think - but could be wrong - that its ususal in companies going public and attracting "big hungry" investors that cross the line
Convenient suicides, like Boeing whistleblower
Sometimes when you do the right thing your conscience just drives you mad. /s
They just activate Covid Vaccine
Could also be literally someone like Musk to make it look exacly this way.. Never think you've seen through if its public info
He probably contributed to copyright infringement while working there or saw negative effects in family and friends of what he was helping build.
Another issue is that engagement algos can be very reflective of what someone is saying, so an accuser can get the feeling that the internet is accusing him back.
@2beJT Yeah, no. Not everything is so cut and dry as what you might initially think. It's pretty easy to call any type of takeout as self takeout. These people do not play by the same rules.
Zuck and Elon vs OpenAI? All I can say is that there will be no victory for the average person there.
tf elon is going to do jump around and cry?
At least Elon understands what humankind need to thrive the other guys doesn’t care
Don't forget China. Smaller and somewhat 'open' models with almost state-of-the-art performance are all the rage right now.
Daddy Musk is just mad he doesn't have more stake in OpenAI now that it's making money.
You don't know what you're talking about. Zuckerberg has open-sourced the llama LLM's.
why whisterblowers die?
its always a accident. unlucky.
He told the truth.
bc they can't keep their mouth shut..
Karen Silkwood was the first WB if I'm not mistaken, there are several podcasts you can listen to about her story
its not a coincidence for sure
Clinton’s
Every fucking time.
The worst part is that this happens so often nowadays that you can never possibly tell if it was an accident or not. The fact that it happens so often... makes you scratch your head.
Move along; nothing to see here
Of course you can.
It was no accident.
Believing anything else is navieté.
if it happens too often, how can it be an accident
Whistleblower hitman
@@yt-sh Exactly
Something to consider that seems to have been omitted about the results shown at 8:35, is that the OpenAI model was allowed 50 submissions per question, while each of the human participants were only allowed a single final submission per question. This was also the case in the Math trials as well, where the OpenAI models were allowed hundreds or even thousands of attempts to solve the given problems.
So while it is still very impressive that these machines are able to solve these problems in a comparatively short timeline when compared to a human, these are still well-structured problems with distinctly defined solutions.
While "in the wild", and especially with non-trivial professional use-cases, AI output frequently requires heavy oversight from a field expert to analyze and verify its validity.
Anyone whose toyed around with text, image, song, and code generation knows that it can require multiple attempts to get the desired output (or in some cases, even anywhere in the ballpark of desired output), and the more complex and technical the task, the higher the level of required expertise and scrutiny is needed to make sure the output is correct (or can be modified to be correct).
The only prediction I have about AI is that a handful of individuals will get insanely rich.
Facts 💯. Always follow the money bro😂
As ever always. 🤝
We've been living in nearly unrestricted capitalism since the 60's ... What's the problem? You don't like your living pod?
and a million of people will be poorer
Not physically, like tangable gold, bit or cash but yes.. In that completely trustworthy and mythical arena called the stock market.
Whistleblowers are the ones who kept humanity alive ❤😢
But they cannot live themselves
You talk about Humanity like you don't have any blame... do you have an iPhone, Apple product, or anything made in a foreign country by slave labor?... think about it... "Humanity" is a "concept", not a REALITY... there are still actual Slave Markets in parts of Africa and the Middle East... there are illegal "slave style" workshops in most Western Countries using illegal migrants and children... Ideology and REALITY are two totally different concepts...
The frequency we talk about ‘leaks’ and ‘whistleblowers’ with a company named ‘Open’AI is 2024.
WhistleLeaks AI
Was founded to be open sourced until they realized how much money could be made
@@thecrackin-u8pI don’t even think making money is Altman’s goal. He has become one of those movie supervillains who just wants to see how further he can push this.
@@ScaryYokai I should add power..actually power comes before money
@ScaryYokai but rest assured very very bad people will be using true ai, (kinda are now but ai isnt their yet, talking full Ai, when its teaching itself) not this artsy shit....in 20 years it's gonna be interesting times.....
are they "stealing" content from the internet? Yes. That's the answer... Yes. Let's stop deluding ourselves as if original data is being created. It is not.
It's more a question of what loophole in copyright and/or ToS they found, with a few backroom deals between themselves and websites probably being kept quiet.
Well... synthetic "inspired" data is being tested out as a supplement to training data or as a full replacement, but the research won't lead to anything since synthetic data needs human curation to be usable, which defeats its whole purpose. So the argument will hold for as long as generative AI exists
They aren't. The AI uses data from anything the user provides regardless of their copyright.
The Indian guy committed suicide because his stupidity got called out. ChatGPT is a tool. Whether a user is responsible is none of their business because that is impossible to control.
Ok. Prove it.
If you ask ChatGPT to write a poem about swimming cars made of macaroni, where is that stolen from?
Imagine being a worker at OpenAI and having even a 1% doubt in your head that the company you're working for might have killed an ex-colleague?
Not exactly a great work environment 😮
Welcome to the "future".
Yeah and its bullshit.
New surfs on a land
Most people are fools. They convince themselves to be more perfect slaves.
The money is too good.😢
OpenAI wants to be regulated by the government so the regulations stop/crush it's competition. Lots of ai startups right now - will be interesting to see how it plays out.
Exactly. These corporations have met the threshold that any regulation on them only hurts us exactly like how copywrite laws have done. Once you amass so much information and wealth on something it's hard to take that away from them even with regulations. Disney still has the back catalogs of like 80 years of drawings and countless of scripts.
This is exactly what tobacco companies did in the USA back in the day
I think Altman is the devil. He looks like G-Man from Half-Life
@@timmyhenrikssons haha
We all need to give Dagogo more praise for this amazing channel. Thanks for staying true to the vision.
True. I still share very old videos to educate people (like the Samsung one).
Dagogo? Where is he from?
@@AndresOssahe’s has a very high production quality since the inception of the channel. One of those evergreen types
@@williamowens2063 if I'm not mistaken he's nigerian and lives in Australia
Exactly!
Sure... why do they even think people would believe that another whistleblower ended their live? Really? As if...
he was feeling down, and stuff. the company got "lucky" 🤣
They care to stop an investigation and media coverage not what people think unfortunately.
You should ask yourself that question.
Maybe because there is a long track history of people believing and even those who don't believe eventually forgetting and moving on. Unless you are rich or famous most people don't care.
Been rooting for this turnaround in opinion ngl, I mean these guys literally want the eradication of art as human expression its hard to get more evil and soulless than that.
Inspiration is not allowe...
Sounds like they could be "Too Bad To Succeed". The pursuit to dominate a tech space can skew a companies moral compass in the wrong direction.
Neoliberalism ruins it every time. Pick a single problem in American business/politics/economics/etc. profit before people guarantees the worst outcome
You will have to be willing to walk over bodies to get on top of a high mountain. A rule that is valid for all circumstances.
@@roscojenkins7451 lmao what
@@NewMitchell-wh3fj You don't understand AI word smoothy? Just take a buzzword, throw in some hyperbole, some words from a glossary, blend, and top with some more hyperbole. What it lacks in nutrition it makes up for with ... stuff.
"Too big to fail" just means too big to manage.
No it doesn't lol. It means socialism for the rich, capital oppression for the poor.
No. Actually it means that considerable pay-offs, bribes, and political influence has been so integrated into our financial system that many businesses will fail as a result of one failue.
AI scoring high in tests is because it have access to all the necessary repository... is the same as a person doing these exams with a laptop and internet access to search for things...
Yeah, AI evangelists pointing to these tests when they're like handpicked tests that the AI particularly excel at. The problem is almost none of these actually reflect real world usage
@@FunkyToe369 He said in video that no solutions are publically available and these are phd level questions.
We won't reach AGI until truly logical models surpass us. Generative AI can only be as good as the data it's given, which can't be higher quality than the data us humans have produced
@@rishitkhanna336the information needed to come to a conclusion is still available though. there may not be outright solutions but they still have a lot of info to work with.
Ai corporations are directly profiting from everyone’s work without even paying them anything or asking creators all while convincing the layman to think ‘oh well it’s the future bruh adapt or die.’ Nah, it’s pretty much theft.
Frank Herbert tried to warn us.
Alternative title: The War of the Sociopaths
That whistleblower ending himself is as likely as Epstein or the Boeing whistleblower.
I have been watching ur channel for about 3 years, your tsmc/invida video changed my life financially. I cannot thank you enough dagogo! Keep going cuz!
I haven't seen the video,
What changed man?
@bryantony9836 well it was back before invidia's most recent surge . The video goes into invidias partnership with semicondu tor manufacturer in Taiwan TSMC .
@@terenceking2252 so did you like buy stocks of tsmc? Or nvidia?
@saadderaiya both . Tsmc was flat already and I was already in invidia , but that video convinced me to close some other positions and get heavy into invidia . I'm not super investor savvy or nothing but in. Idia already was showing impressive growth pre and post pandemic so it made alot of sense when tsmc started using invidia's tech along with the AI boom. Sorry for long response.
@ thats really great to hear brother !! How much wear you able to make with that?
OpenAI went full Boeing quite fast. Though main problem is, that all these companies are breed on awesome, capital intensive and hype prone US market - in tthe same country, that has Congressional Hearings for Big Tech but also on Aliens / UFOs.
Yea those hearings about not UFOs, that word is dead, but like Anonamlyous Flying Entity AFE? will be interesting, specially while Trump is in office, he could reveal long hold secrets (which sure exist but idk to what extend) 👀
From the firing of and rehiring of their current CEO Sam Altman to the death of death of former employee and whistleblower Suchir Balaji... THE WORSE IS YET TO COME
A note with o3. While it wasn't trained on the ARC dataset, it isn't default o3 either, but a version fine tuned on the ARC public examples. They also spent significant compute resources estimate to cost $1-10k per task.
There should be an act to protect all whistleblowers worldwide.
one of the best channels i stumble upon long time ago and since then watched all videos
Feels like the old days during the browser wars. One company creates a new feature - the rest copy it and/or improve upon it. Pretty soon the market is reduced to just 2 players - just like everything else in the world.
"Growth for the sake of growth is the ideology of the cancer cell." -Edward Abbey
We humans have always wanted to scape "God" and at the same time "eat it".
And that "ideology" is not only the one of cancer, but what we call "life".
Absolutely no one going to court is going to off themselves. Depressed people don't Whistle-blow and go to court. They just don't.
0:37. Why would you self exit the playable space, when you're in the middle of a PvP battle? Sounds like bull to me.
Because you know you can't beat the boss & you became a sentient NPC probably planning on respwaning at a checkpoint
@@infostations I think you took what I was saying too literally.
OpenAI’s most recent products have absolutely not received a Lukewarm yawn from its audience. O1 has been completely amazing.
Search, Voice Mode, and other releases made during the twelve days of Christmas campaign were impressive too.
Please, which software do you use to create your videos
BidenpartyAIdotgay
microsoft powerpoint
windows movie maker
My guess would be premiere pro, or its possible he doesn’t want to pay a yearly fee and uses Final Cut Pro for Mac users or DaVinci resolve.
Hand stitching film
I think the whole legal stuff regarding copyright and training of models is going to be a massive slam on the brakes of all 'above ground' models. Big platforms will probably exclusively license their content for AI training and it will probably cause publicly available models to stop being trained until most of the internet-pie has been sliced. Or they have to figure out a way to mark a source of the trained data so that they can enable/disable it based on shifting licensing agreements. But I'd guess platforms and web application firewalls will soon block or log (for litigation) AI training requests from parties not currently licensed.
The transformation from a non profit to a profitable company is outrageous
Dagogo you are a national treasure. Thank you for all your hard work.
it's really crazy how nobody is talking about the book the elite society's money manifestation
Scam book
AI: "We're selling you regurgitated data that you unknowingly provided us"
happy new year, Dagogo!
wish you a great 2025!
Cheers Fella ,Happy New To You And The Family 👍
Amazing channel mate
You work for 40yrs to have $1M in your retirement, meanwhile some people are putting just $10K into trading from just few months ago and now they are multimillionaires
As a beginner what do I need to do? How can I invest, on which platform? If you know any please share.
Wow that's huge, how do you earn that much?
I'm 37 years old and I've been looking for ways to be successful, please how??
Yeah, 253k from Maureen duke, looking up to acquire a new House, blessings.
I thank Maureen Duke who has always been there to help me with detailed analysis and recommendations that I would not have had access to otherwise.
I am amazed that you mentioned and recommended Ksenia Alice. I came across one of her client's testimonial on CNBC news last week.
Your channel is very interesting and helps me keep updated with tech and economy. Thank you for all the hard work and a happy new year!
Compretition is good, it always brings forth innovations.
We've got Openai mafia before gta 6 🗣️🔥🔥🔥
I can't wait for GTA6 come out so these jokes stop forever
@mattcy6591 never
😂😂😂😂
Thank you for not using sloppy AI in your videos❤
Another sign of times getting tough, sad and bad.
First, videos are great as always. A small detail at 8:23 : o3 was only revealed as you previously said, not released. Also and on a side note, love the music and “dimmed radio voice” at the beginning and ending, it both feels dream-like and nostalgic.
I love the way ColdFusion explains the things. The most important, the information is unbiased which is very difficult to find elsewhere. Keep up the work! Wish a lot of growth for ColdFusion in the year 2025! 😄💫
Just came for the intro and outro song
Dagogo thank you for this excellent report on Open AI. We will see how this develops but it isn't looking that good for the moment.
As always detailed journalism with nutral perspective ❤
7:20 - 8:00 what is the song used here? So many song on burn water channel I can’t find it
If there isn't a criminal investigation, it would be a disgrace and injustice to this whistle blower.
OpenAI is expanding too fast with not enough control and governance 😢.
Great video
the dude is in all ways right these AI uses copyrighted materials the problem is proving it does since its meshed with other stolen materials from other stolen materials its hard to distinguish
In the AI market, the first thing that happens to any obstacle would be "suicide."
I love the videos. I learn something every time
This is kinda like when the CIA does some horrendous operation and then the operation is declassified 20 years later. The public goes 'wtf!?' the CIA is like 'what u gonna do bout it cuh?'
With all these abnormal murders its like they want to get used to them and live life care free
“Too big to fail” is the defining statement of what the failure of life liberty and pursuit of happiness means.
Love your music choices
There should have been a greater focus on the O series of models in the video. It was the defining breakthrough of 2025 and a much bigger deal than any of the drama. Many who report on AI seem to focus more on the drama than the technology itself.
To keep up with the most recent developments in the technology itself, and their broader implications explained in simple terms, I would highly recommend the channel, Ai Explained.
You are very brave for reporting this.
Good video, Macca
People are jealous OpenAi wants to profit from their own work.
0:50 Executives are paid very highly and would rarely resign such a high paying position. So it's a red flag that the fact top executives(so many of them) are resigning. It means they don't feel safe in their positions, most likely as they are seeing very illegal and unethical practoces that isn't even worth the risk of being sued upon. The fact these people are resigning willingly means there's something sinister going on in OpenAi.
I wouldn't include ChatGPT in the "AI fatigue" category. Image and video production, sure... maybe... (a few communities that I belong to would beg to differ) , but CharGPT has manged to give answers quickly to questions I have had for years that vigorous web searches had not provided.
here for the ride
Thanks for sharing it. ✌️
Small correction. It was shown they used the scoring tests in their training data.
hello non american 🇺🇸
open ai altman vs musk is a movie 🍿 script writers dream. *PLUS* that person gets to use chatgpt, so that writer will have no problem, just gotta watch 👀 to let it unfold.
movies have been rubbish for 20 years. if you still watch, i feel sorry for you.
10:40 funny how this always happens when it's about a whistleblower. Boeing
I just saw an article that said OpenAI's latest model costs $1000 PER QUERY . That's how much computational power they're using. It's unimaginably unsustainable. The ONLY reason openAI hasn't collapsed is because they have a $65 Billion / Quarter cash pile coming from Microsoft.
O3> is beyond what competitors can achieve at the moment
By the time they get close, OpenAi would probably be on o5
They'll be fine
o4 or o5
DeepSeek is coming close
@aaronalquiza9680 I also use deepseek/deepthink, it's really impressive, but nowhere close to o1/o1pro. Ofc I must consider the fact that deepseek is practically free or very low in cost. With o3 mini coming in Feb/March and the full o3 coming soon after, the gap is very large
How can you know? As far as they announced, O3 is pretty much unusable by anyone. Look at the cost of the thing.
AI is Sooooo good. its probably why Cursor decide out of nowhere to delete entire features from an existing working software. When asking why it can't answer anything logical. We now need to get trained on the pitfalls of AI.
We are in the age of running LLM at home, there are already several LLM brands that are not too user friendly to install, once this aspect is cleaned up, this will take off like a rocket.
As far as I know it'll still take at least 2-3 years until we'll be able to run full LLMs like Sonnet 3.5 and GPT4 entirely on device
@@sirsamiboi we can run LLM like ChatGPT in our home. LLAMA can compete with it but it requires at least 64 GB RAM and a strong GPU last time I checked. There are also smaller models but they are not satisfying as a chatbot.
Ollama2 installs in minutes, requires no accounts to use, runs on Windows, Linux, or Mac. Everything sets up effortlessly, The most user-unfriendly thing is you have to run it in a command prompt. It's very powerful and runs on your own local hardware. I've run it on a thinkpad laptop with no GPU and it will answer queries, sometimes it will take a while to get replies (1 to 2 minutes). On a machine with a modern GPU, queries return instantly always.
Absolutely no need to step down to using something like OpenAI's products, etc. and support companies I despise, when I can run it locally, it doesn't even need an internet connection to function and get you answers.
"Blown out of the park"? - 'Blown out of the water' (WWII) 'Knocked out of the park' (Baseball) - Like this, eventually, NOTHING will matter.
Owsome video, keep going mate!
6:42 “non-profit” is the biggest lie of 21st century. Somehow anyone around non-profit always becomes staggeringly entitled rich person
AI fatigue? Is OpenAI over? Seriously? I usually enjoy your channel but this is an insane amount of drama and hyperbole.
Your channel is a real treat, You need to cover the biased mainstream media coverage of the Gaza genocide
how do you know, that similar questions to the arc benchmark are not in the training data?
I immediately thought about copyright infringement when I heard about AI, but nothing was addressed...
copilot is OpenAI. Any revenue from Microsoft copilot is a win for OpenAI. Microsoft inadvertently owns OpenAI due to their extraordinary partnership.
Liked this one Dagogo, even though I have my doubts in ethical AI implementation. 👍💪✌
Tax the AI companies now. Massively. That way as a government you'll have more money available to give benefits to those that have lost their jobs to AI in the next 1-3 years and will be claiming under joblessness. It's guaranteed the majority of people in offices right now will lose their jobs to software devs and IT guys maintaining the AI for the companies. They're aiming AI at the housing market which has disaster written all over it. EVERY SINGLE instance of money-saving techniques in history introduced to an industry have been taken on and applied. EVERY single one, every time. Quicker techniques, cheaper processes, whatever it is. "AI can save you money by cutting X% of your workforce, and you get the workforce to teach it their jobs before they leave".
I'm still waiting for AI to develop internal code standards to be transparent and identify itself. Meaning, AI should be capable of interrogating another product (audio, video, text, image, etc) and confirm if it is original or if it is AI generated. This TRUTH should be encoded in the product geocode.
Any model that can reliably detect generated content can be used to train another model that would fool the detector, it's called a Generative Adversarial Network. A perfect "AI detector" can always be used to fool itself
@@landlubbber thanks for your AI response.
How about Cold Fusion doing a series of videos explaining the Swiss political system of real democracy to the World
*Aaron Swartz* passing was after he did what ..?
Now surpassed by Google, X, Antropic and a Chinese company on the lmsys leaderboard. I wonder when the customers will move too.
Quite often the originals don't usually win. It is usually the ones who sit back and pick the scraps that win. Opening doors for the one who will sit in the car...
I'm a bit conflicted about copyright law being applied in the same way when used for training data. Maybe they should pay something for using it to train models, but I don't think it should be classed the same licensing content on a platform like Netflix for example, as you're not distributing the content itself, it's more equivalent to how human's learn from various sources, if a person learns something from an article online and goes on to share that information is that breaking copyright law? I feel copyright law would need updating to be suitable for the changes in modern times.
The artists don't wanna accept that though, I think before any progress can be made people need educating.. people need a reality check on where they draw influence from..
I think people want to cling onto the fact that we are more special than a computer program.
Every argument you see from artists against this use really undefinable language like "soul" "creativity" etc etc.
This ai tech is heading down the path of how our brains work.. it's miles of yet but the principles are already there..
The models are just super primitive at the moment.
I understand the conflict man.These things are very complex. But it's not equivalent as to how humans learn, and this is a dangerous idea. Our leaning process has evolved from millions of years of "information processing", it involves many parts of a body. It's not just the brain. We are not regurgitating statistics like the machines. And we develop style. It's incredible that these LLM's with all the art they consumed can't develop and original, never seen before style.
LLM's don't even know what they are doing... And, dude, just why we keep using the human analogy? We should have more rights than machines, more rights that corporations. They are stealing. Let's call it like it is: A theft. They are stealing from one of the most vulnerable people of our societies and by doing so they are depriving them of their jobs too. They employ slave labour as well. I have no words for how shitty this thing is and this is just the beginning! At least Google and many other tech companies took some time before they were the evil ones. These guys are bad form the start.
So I don’t wanna burst your bubble here or anything but my great uncle worked with arpa before it became darpa and he said they were working on LLm’s back in the 1970’s and that our thaad and aegis defensive missile systems were being incorporated back before they were known systems. He said imagine the most advanced thing possible for technology and then go 40 years into the future and that’s what darpa is working on currently.
Those “UFO’s” you hear about on the news etc, that’s darpa physically testing systems. Haarp systems pushes extra energy to the crafts when needed for high maneuvering amongst many other things…
The world is gonna be just fine, we’ve already given it over to the AI masters, it just hasn’t been allowed to be known officially yet…
Your uncle was a duck. Quack quack quack...
I have heard similar statements from people who used to work on black projects. I wonder if there are certain problems with these systems that make widespread adoption not feasible?
Bought XAI303e after watching your video and others, super excited! 💰
Brilliant as always
Didnt mention OpenAI involvement in autonomous weapon platforms
This just makes me more bullish on them. When you want to know who's the best you simply need to ask who's getting attacked by everyone from almost all sides for complete bs reasons?
Great video.
You mentioned "breathing microplastics"... do you know that you are breathing microplastics every time you enter a hot vehicle in the summertime?... that "film" on the inside of your windshield is outgassing plastics from everything plastic inside your vehicle... especially the dashboard... a majority of the dust in a carpeted room is probably microplastics... Those "reusable/refillable" home soda machine bottles have, in small print, an "implied WARNING" by listing and "EXPIRATION DATE" on each plastic bottle... WHY?... Microplastic Deterioration... IT'S ALL AROUND US...