San Francisco – Sam Altman, CEO of OpenAi, does not hide how much your company plans to spend on your search to build artificial intelligence.
“You should expect OpenAi to spend trillions of dollars on things like data centers building in the not too distant future,” Altman said recently, referring to the huge computing facilities that feed the company’s AI technologies.
“You should expect a lot of economists to worry and say ‘this is so crazy. It’s so reckless’ or something. And we just say, ‘Do you know one thing? Let us do our work.’
Free tool
XP Simulator
Learn in 1 minute how much your money can yield
So what exactly is this work? While the technology industry spends and spends, turning agricultural land into data centers and IA researchers into some of the country’s highest paid workers, it has been difficult to explain what it is building and why it is spending so much money.
Are they building a smart AI system as humans? A divine machine that will change the world if it does not destroy humanity first? Are they working on more sophisticated versions of software they sell for decades? Is all this money going to a bold plan to create fake virtual friends and more effective ads? Or are they just afraid of being left behind in what everyone is doing?
Here is a summary of the visions, from the most plausible to the most fantastic, and why they are pursuing these ideas:
Continues after advertising
The promise: a better search engine
Chatbots work a lot as a search engine, except that they generate simple English answers instead of a list of blue links. This can be a faster, easier and intuitive way to answer questions, although chatbots often make mistakes and even invent things.
Why are they building this?
Google’s search engine is the most lucrative business in the technology industry. If companies could offer a better way to seek information, they could capture a billion market.
Continues after advertising
How close are it from reality?
Hundreds of millions of people already use chatbots for information. ChatgPT alone has over 700 million users per month.
But profit from this technology is a challenge. Operating a chatbot is significantly more expensive than serving a common site. And technology does not necessarily fit the traditional model of making money from search engines: digital advertising.
Continues after advertising
OpenAi sells a version of chatgpt for $ 20 a month, and according to the company, at least it covers the delivery cost. But these subscribers represent less than 6% of ChatgPT users.
The free version is still in the red, as OpenAi has not yet started trying on ads. Google, on the other hand, generates $ 54 billion in ad revenue every quarter with its search engine, used by about 2 billion people a day.
(New York Times sued OpenAi and Microsoft, alleging copyright violations of AI -related journalistic content. Companies denied these allegations.)
Continues after advertising
The promise: tools that make office workers more productive (and perhaps replace them)
The technology that feeds chatgpt not only answers questions, but it is a tool that can help people do their jobs. IA can generate computer programs, summarize documents and meetings, write emails and even use other software, such as online spreadsheets and calendars.
Why are they building this?
Technology executives believe that AI can turn the business world by entering law firms, hospitals, newsrooms and more. Companies like Microsoft and OpenAi are already generating significant revenue from selling AI systems that generate computer programs.
Continues after advertising
Amazon, Google, Meta, Microsoft and OpenAi plan to spend more than $ 325 billion combined on gigantic data centers this year. This is 100 billion more than Belgium’s annual budget. Over time, about 10% of the infrastructure will be used to build AI technologies, while 80% to 90% will be used to deliver these technologies to customers, according to Amazon CEO Andy Jassy.
How close are it from reality?
Many companies are already testing AI. But great implementations in the US economy do not happen yet. Unless companies like Amazon, Google and OpenAi continue to improve these technologies, adoption can be slower than expected.
Almost 8 out of every 10 companies started using Ia Generative, but as many as they said that “there is no significant impact on the end result,” according to McKinsey & Co. survey.
“The card house will start crumbling,” said Sasha Luccioni, a researcher at the Ai Hugging Face startup. “The amount of money being spent is not proportional to the money that is entering.”
The Promise: An Assistant for Everything
Technology companies are also incorporating similar chatbots technology into a wide range of consumer products and services. AI, they say, will function as a digital assistant that appears where necessary.
The goal is adding technology to their smart glasses, allowing people to identify sights as they walk and translate street signs in foreign countries. Amazon sees AI as a way to improve everything from its shopping sites to its voice assistant Alexa.
Why are they building this?
If you start using a digital assistant, the company behind the bot has more ways to capture your attention and, in the end, sell things to you.
So these companies are adding AI technology to as many online devices and services as they can, hoping to control the way you use the internet.
“Everything will be transformed with AI,” said Rohit Prasad, Senior Vice President of Amazon. “This is not a scientific project.”
How close are it from reality?
AI goggles of the goal are still a niche product worn by a few million people. Alexa da Amazon is much more popular, but its audience is still small compared to all computers and phones in the world.
Alone, Alexa has been damaging since its launch for over a decade. It is mainly used to boost other products and services.
When Amazon updated Alexa with new AI technology, he gave the update for free to those who pay for the Prime program. AI can make it more popular, but it is unlikely to become profitable soon.
The Promise: Friends of IA
Goal and several startups, including Character.ai and Elon Musk’s Xai, are beginning to offer AI bots that provide a new type of company. People can interact with these bots on social networks just as they interact with friends.
“The average person wants more connectivity, connection, than he has,” said Mark Zuckerberg, CEO of Meta, in a recent interview.
Why are they building this?
Zuckerberg and Musk command social networks and could charge for these virtual friends. Musk offers his bots through a subscription service that costs $ 300 a month.
The goal could also charge a signature for virtual friends, just as OpenAi makes for chatgpt, although the goal preferred to increase ad recipe keeping people on websites like Facebook, Instagram and WhatsApp. (The company also applies in this area. Recently, the goal found that people have almost 7% more chance to click on ads created with new AI techniques.)
How close are it from reality?
Some people already treat chatbots as friends. But Ia’s company is starting to receive heavy criticism. These technologies can drive people away from human relationships and lead them to alarmingly delirious behaviors.
It’s still for years to become a viable market, and it’s just one of the many scenarios that companies are exploring.
Some observers compare what technology executives are making pieces move in a chess board – trying to beat their rivals in the next great technology.
“So much can reside in so few people,” said David Cahn, a partner of the risk capital firm at Silicon Valley Sequoia, “and they are playing a chess game that has implications for all of us.”
The Promise: Scientific Advances
Dario Amodei, CEO of Anthropic, one of OpenAi’s main rivals, believes that in a few years – perhaps next year – artificial intelligence will be like having a “country of geniuses on a data center” that can work together to solve the biggest scientific problems our society faces.
Why are they building this?
Technologists as I amodeed believe this type of technology will change life as we know. Last year, in a 14,000 -word rehearsal, he said AI could eventually heal cancer, end poverty and even bring world peace.
He predicted that in a decade the AI would double the average life expectancy for 150 years.
How close are it from reality?
It is unclear how these technologies will be built – or if they are really possible.
But Google’s senior senior vice president, laboratories, technology and society, said that while Google pursues more ambitious goals will create technologies that can be used immediately. As an example, he cites Alphafold, a Google -developed system that helps speed up the discovery of medicines in a small but important way, and recently won the Nobel Prize in Chemistry.
A Google spin-off called Isomorphic Labs intends to make money by helping pharmaceutical companies use this type of technology.
The Promise: It was as intelligent as a human, or more
Executives like Zuckerberg and Demis Hassabis, head of Google’s Deepmind Research Laboratory, say their companies are pursuing general artificial intelligence, or AGI, abbreviation for a machine that can match the powers of the human brain, or an even more powerful technology called superintelligence.
Why are they building this?
Many technologists are determined to pursue the greatest goal they can imagine: the superintelligence. They have been looking for this dream since the 1950s.
How close are it from reality?
Terms like Ages and Superintelligence are difficult to define. Scientists do not even agree on a definition for human intelligence.
But a machine that really equals the powers of the human brain has been for many years, perhaps decades or more.
No one explained how companies will profit from this type of technology. While technology companies spend hundreds of billions on new data centers, they are making a leap of faith.
This leap is fed by the same mixture that often moves the silicon valley tycoons, said Oren Etzioni, founding CEO of the Allen Institute for Ai: greed, ego and fear of being overcome by an unexpected discovery. “If I had to answer a word,” said Etzioni, “it would be fomo.”
The fear of being out is not cheap. Altman said that while he and his rivals pursue these high goals, some investors may be spending too much. Researchers can develop ways to build IA using much less hardware. People may not want AI technologies that these companies are building. The rapid improvement of AI technologies in recent years can slow down or even hit a wall. The entire economy can change for unbleed reasons.
“Some of our competitors will fail and some will do very well, and that’s how capitalism works,” said Altman. “I suspect someone will lose a phenomenal amount of money.”
c.2025 The New York Times Company
