JOHN YANG: OpenAI, the company behind ChatGPT, is one of the world's most famous and secretive companies.

It's on a mission to try to develop artificial general intelligence, or AGI.

That's a theoretical type of AI that possesses human intelligence and can perform any intellectual task a person can, as opposed to more specific tasks like image recognition or how to win a chess.

Investigative journalist Karen Hao's new book, "Empire of AI: Dreams and Nightmares, and Sam Altman's Open Air," offers a skeptical look at where the company is headed.

Here's Ali Rogin.

ALI ROGIN: Karen, thank you so much for joining us.

You've been following OpenAI from the very beginning when it introduced itself as a research nonprofit with very lofty goals.

Tell us about those early days.

KAREN HAO, Author, "Empire of AI": Yeah, so I was the first journalist to ever profile OpenAI and I embedded within the company for three days in August of 2019.

And the reason why I thought it was worth profiling them in that moment in time is because they were beginning to shift away from those nonprofit origins.

They had just restructured to put a for profit within the nonprofit that suggested there would be some kind of commercial intent later on.

And I was asking executives and employees to explain to me what was the mission.

And they insisted that the mission was the same, that they were trying to ultimately build AI completely without commercial interest, and they were going to be open and collaborative and they wanted this technology to be ushered into the public purely for social benefit.

And that is not what I found.

I found that they were very competitive, they were very secretive, and that certainly they had to have some kind of commercialization plan because they had just received a $1 billion investment from Microsoft.

ALI ROGIN: And of course, we can't talk about the company and the ethos of the company without talking about its CEO, Sam Altman.

You track his rise in the company and then his firing and then subsequent reinstatement, which really shocked a lot of folks in Silicon Valley.

How much has Sam Altman and his views informed the direction of AI in Silicon Valley?

KAREN HAO: I really think OpenAI is a manifestation of Altman.

He is a product of Silicon Valley himself and he is very much a product of this philosophy of growth at all costs.

Try and be ambitious about changing the world and assume that the world is a zero sum game and that winners take all in this race.

And so you see a lot of how Open Air operates today is an embodiment of that.

They chose to take an AI development approach that is based on this growth at all costs mentality, where they're trying to train their models on the entire internet.

They're trying to train them on massive supercomputers that are starting to really strain the energy resource availability around the world.

They are trying to exploit a lot of labor to fortify that particular AI development paradigm.

And ultimately he is positioning AG development as this aggressive race where OpenAI needs to be number one.

Because if not, if an American company is not going to beat that global race, then China will overtake it.

And there is this zero sum game mentality around the rhetoric.

ALI ROGIN: You have a position in this book, and it is that AI evolving in this way is indeed a threat to humanity.

What do you see as the biggest risks?

KAREN HAO: For me, it is very different from the doomer threats where the doomers believe that AI could fundamentally go rogue and it could develop consciousness and therefore kill everyone.

That is not what I think is actually the risk here.

I do not think that I will go rogue.

There is no scientific evidence for that particular claim.

But what we are seeing in the real world right now is that this particular AI paradigm is creating an enormous consolidation of economic and political power in the hands of these companies.

And my title, "Empire of AI" is a nod to the argument that these companies actually need to be thought of as new forms of empire because they are laying claim to extraordinary historic levels of resources.

They are exploiting an extraordinary amount of labor, they are monopolizing knowledge production.

So most of the AI research is actually filtered through the lens of what is good for these companies versus not because they have a monopoly on the top AI talent in the world.

And ultimately all empires have this narrative that there are evil empires and there are good empires.

And the reason why they, the good empire, have to be an empire in the first place and engage in all this resource and power consolidation is because they must be strong to beat back that evil empire.

But ultimately we are just seeing a profound, profound unprecedented consolidation of power.

And the threat is when we reach an age where people at the top can just do whatever they want, develop this technology however they want, deploy it however they want, and potentially lead to mass environmental consequences, mass economic consequences, huge amounts of job loss, that ultimately most of the people in the world are going to feel this loss of agency in determining their own future.

And when people do not feel that they have a voice to self-determine, democracy dies.

So the greatest threat is to democracy.

ALI ROGIN: You also document many of the problems with AI and the unfettered growth that it is experiencing.

Proponents, including Sam Altman himself, would argue that there are dramatic tremendous benefits as well.

Is there anything good you see about AI and the pursuit of artificial general intelligence?

KARON HAO: AI is such an interesting word because it's sort of like the word transportation and that you have bicycles, you have gas guzzling trucks, you have rocket ships, they're all forms of transportation, but they all serve different purposes and they have different cost benefit trade-offs.

And to me the quest to artificial general intelligence has the worst trade-offs because you are trying to build fundamentally an everything machine, but ultimately it can't actually do all of the things.

So not only do you confuse the public about what you can actually do with these technologies, which leads to harm because then people start asking it for things like medical information and instead getting medical misinformation back.

But also it requires all of these things that I described, the colossal resource consumption, the colossal labor exploitation.

But there are many, many different types of AI technologies that I think are hugely beneficial.

And this is task specific models that are meant to target solving a specific well scoped challenge, something like integrating renewable energy into the grid, weather prediction, drug discovery, health care, where you identify cancer earlier on in an MRI scan.

These are all very task specific.

It's very clear what the use case is.

It's -- you can curate very, very small data sets, train them on very, very small computers.

And I think if we want broad based benefit from AI, we need broad based distribution of these types of AI technologies across all different industries.

ALI ROGIN: The book is "Empire of AI."

Karen Hao, thank you so much for bringing your insights.

KAREN HAO: Thank you so much Ali.

JOHN YANG: Sam Altman declined an interview request.

We remain open to it in the future.