Meet GPT-3. It Has Learned to Code (and Blog and Argue).

Alex Eben Meyer | The New York Times
Alex Eben Meyer | The New York Times
TT

Meet GPT-3. It Has Learned to Code (and Blog and Argue).

Alex Eben Meyer | The New York Times
Alex Eben Meyer | The New York Times

This summer, an artificial intelligence lab in San Francisco called OpenAI unveiled a technology several months in the making. This new system, GPT-3, had spent those months learning the ins and outs of natural language by analyzing thousands of digital books, the length and breadth of Wikipedia, and nearly a trillion words posted to blogs, social media, and the rest of the internet.

Mckay Wrigley, a 23-year-old computer programmer from Salt Lake City, was one of the few invited to tinker with the system, which uses everything it has learned from that vast sea of digital text to generate new language on its own. Mr. Wrigley wondered if it could imitate public figures — write like them, perhaps even chat like them.

One of his experiments involved a pop psychologist, Scott Barry Kaufman. The system took in Mr. Kaufman’s name and a topic for discussion: creativity. Then, when asked “How do we become more creative?” GPT-3 responded instantly:

I think creative expression is a natural byproduct of growing up in a diverse world. The more diverse the world is, the more you get exposed to different people, to different opportunities, to different places, and to different challenges. And the more diverse that is, the more likely you’ll be to be able to put the dots together to form something new. And in many ways, I think if you want to be creative, you have to go for it. If you want to be a writer, you have to write, if you want to be a musician, you have to create music, if you want to be a comedian, you have to create comedy, and the more you create, the more likely it is that you’ll stumble onto some interesting stuff.

Later, when Mr. Wrigley posted the paragraph on Twitter, somebody looped in the real Scott Barry Kaufman. He was stunned. “It definitely sounds like something I would say,” the real Mr. Kaufman tweeted, later adding, “Crazy accurate A.I.”

In the weeks since its arrival, GPT-3 has spawned dozens of other experiments that raise the eyebrows in much the same way. It generates tweets, pens poetry, summarizes emails, answers trivia questions, translates languages, and even writes its own computer programs, all with very little prompting. Some of these skills caught even the experts off guard.

For many artificial intelligence researchers, it is an unexpected step toward machines that can understand the vagaries of human language — and perhaps even tackle other human skills.

“It is surprising to me, and to a lot of people,” said Melanie Mitchell, an A.I. researcher at the Santa Fe Institute, an independent lab in New Mexico, who is among those experimenting with the system. “It is hard to figure out exactly what this thing is doing.”

GPT-3 is far from flawless. It often spews biased and toxic language. And if you ask for 10 paragraphs in the style of Scott Barry Kaufman, it might give you five that are convincing — and five others that are not. When Mr. Kaufman actually chatted with his A.I. doppelgänger, he was not nearly as impressed as he was at first:

Scott Barry Kaufman: What is imagination?

“Scott Barry Kaufman”: What is imagination? You should tell me!

Scott Barry Kaufman: You should know!

“Scott Barry Kaufman”: Nope, you should know. You are the one who knows what imagination is. What do you think it is?

But even this looked like real conversation. And notably, it showed what the system could do entirely on its own, without extra coding or fine-tuning.

GPT-3 is the culmination of several years of work inside the world’s leading artificial intelligence labs, including OpenAI, an independent organization backed by $1 billion dollars in funding from Microsoft, as well as labs at Google and Facebook. At Google, a similar system helps answer queries on the company’s search engine.

These systems — known as universal language models — can help power a wide range of tools, like services that automatically summarize news articles and “chatbots” designed for online conversation. So far, their impact on real-world technology has been small. But GPT-3 — which learned from a far larger collection of online text than previous systems — opens the door to a wide range of new possibilities, such as software that can speed the development of new smartphone apps, or chatbots that can converse in far more human ways than past technologies.

As software designers, entrepreneurs, pundits and artists explore this system, each new experiment stokes an already heated debate over how powerful this breed of technology will ultimately be. While some say it may be a path toward truly intelligent machines, others argue that these experiments, while endlessly fascinating, are also misleading.

“It is very fluent,” said Mark Riedl, a professor and researcher at the Georgia Institute of Technology. “It is very articulate. It is very good at producing reasonable-sounding text. What it does not do, however, is think in advance. It does not plan out what it is going to say. It does not really have a goal.”

An ‘emergent quality’
Jordan Singer is a product designer at Square, the Silicon Valley mobile-payments company. He helps design the company’s smartphone apps, building the graphics, menus, buttons, and other widgets that define an app’s look and feel. When he heard about GPT-3, he wondered if this automated system could do his job.

He fed the system a simple description of a smartphone app, and the computer code needed to create the app. The description was in plain English. The code was built inside Figma, a specialized design tool used by professionals like Mr. Singer.

He did this a few more times, feeding the system several more English-language descriptions alongside the matching Figma code. And when he was done, GPT-3 could write such code on its own.

If he described a simple app for posting and viewing photos as a user would on Instagram, the system generated the code needed to build it. This code was sometimes flawed. But typically, if Mr. Singer made just a tweak or two, it worked as he wanted. “It’s not absolutely perfect,” he said. “But it is very, very close.”

This behavior was entirely new, and it surprised even the designers of GPT-3. They had not built GPT-3 to generate computer code, just as they had not built it to write like Mr. Kaufman or generate tweets or translate languages. They had built it to do just one thing: predict the next word in a sequence of words.

GPT-3 is what artificial intelligence researchers call a neural network, a mathematical system loosely modeled on the web of neurons in the brain. This is the same technology that identifies faces in the photos you post to Facebook and recognizes the commands you bark into your iPhone.

A neural network learns such skills by pinpointing patterns in vast amounts of digital data. By analyzing thousands of cat photos, for instance, it can learn to recognize a cat.

About three years ago, researchers at Google and top labs like OpenAI started designing neural networks that learned from enormous amounts of prose, including unpublished books and Wikipedia articles by the thousands. These universal language models could be applied not just to one task, like translation, but to many.

GPT-3 analyzed digital prose on an unprecedented scale, spending months looking for patterns in huge amounts of text posted to the internet. In this way, it learned to predict the next word in a sequence. If you type a few words into GPT-3, it will keep going, completing your thought with entire paragraphs of text.

But in acquiring this specific skill, it learned much more. During its months of training, GPT-3 identified more than 175 billion parameters — mathematical representations of patterns — in that sea of books, Wikipedia articles, and other online texts. These patterns amount to a map of human language: a mathematical description of the way we piece characters together, whether we are writing blogs or coding software programs. Using this map, GPT-3 can perform all sorts of tasks it was not built to do.

Before asking GPT-3 to generate new text, you can focus it on particular patterns it may have learned during its training, priming the system for certain tasks. You can feed it descriptions of smartphone apps and the matching Figma code. Or you can show it reams of human dialogue. Then, when you start typing, it will complete the sequence in a more specific way. If you prime it with dialogue, for instance, it will start chatting with you.

“It has this emergent quality,” said Dario Amodei, vice president for research at OpenAI. “It has some ability to recognize the pattern that you gave it and complete the story, give another example.”

Previous language models worked in similar ways. But GPT-3 can do things that previous models could not, like write its own computer code. And, perhaps more important, you can prime it for specific tasks using just a few examples, as opposed to the thousands of examples and several hours of additional training required by its predecessors. Researchers call this “few-shot learning,” and they believe GPT-3 is the first real example of what could be a powerful phenomenon.

“It exhibits a capability that no one thought possible,” said Ilya Sutskever, OpenAI’s chief scientist and a key figure in the rise of artificial intelligence technologies over the past decade. “Any layperson can take this model and provide these examples in about five minutes and get useful behavior out of it.”

This is both a blessing and a curse.

Unsafe for work?
OpenAI plans to sell access to GPT-3 via the internet, turning it into a widely used commercial product, and this year it made the system available to a limited number of beta testers through their web browsers. Not long after, Jerome Pesenti, who leads the Facebook A.I. lab, called GPT-3 “unsafe,” pointing to sexist, the racist and otherwise toxic language the system generated when asked to discuss women, Black people, Jews, and the Holocaust.

With systems like GPT-3, the problem is endemic. Everyday language is inherently biased and often hateful, particularly on the internet. Because GPT-3 learns from such language, it, too, can show bias and hate. And because it learns from internet text that associates atheism with the words “cool” and “correct” and that pairs Islam with “terrorism,” GPT-3 does the same thing.

This may be one reason that OpenAI has shared GPT-3 with only a small number of testers. The lab has built filters that warn that toxic language might be coming, but they are merely Band-Aids placed over a problem that no one quite knows how to solve.

“They are doing the right thing by not just publicly releasing GPT-3,” said Allison Koenecke, a Stanford researcher who explores unwanted bias in A.I. systems. “A lot is still up in the air.”

The onus is ultimately on OpenAI to ensure that this behavior remains in check, said Liz O’Sullivan, a vice president with Arthur, a company that helps businesses manage the behavior of artificial intelligence technologies. As it stands, she said, OpenAI is “passing along legal and reputation risk to anyone who might want to use the model in consumer-facing applications.”

Other experts worry that these language models could help spread disinformation across the internet, amping up the kind of online campaigns that may have helped sway the 2016 presidential election. GPT-3 points to a future in which we are even less sure if what we are reading is real or fake. That goes for tweets, online conversations, even long-form prose.

At the end of July, Liam Porr, a student at the University of California, Berkeley, generated several blog posts with GPT-3 and posted them on the internet, where they were read by 26,000 people. Sixty viewers were inspired to subscribe to the blog, and only a few suspected that the posts were written by a machine.

They were not necessarily gullible people. One of the blog posts — which argued that you can increase your productivity if you avoid thinking too much about everything you do — rose to the top of the leader board on Hacker News, a site where seasoned Silicon Valley programmers, engineers and entrepreneurs rate news articles and other online content. (“In order to get something done, maybe we need to think less,” the post begins. “Seems counterintuitive, but I believe sometimes our thoughts can get in the way of the creative process.”)

But as with most experiments involving GPT-3, Mr. Porr’s is not as powerful as it might seem.

The flaws nobody notices
In the mid-1960s, Joseph Weizenbaum, a researcher at the Massachusetts Institute of Technology, built an automated psychotherapist he called ELIZA. Judged from our vantage point in 2020, this chatbot was exceedingly simple.

Unlike GPT-3, ELIZA did not learn from prose. It operated according to a few basic rules defined by its designer. It pretty much repeated whatever you said to it, only in the form of a question. But much to Dr. Weizenbaum’s surprise, many people treated the bot as if it were human, unloading their problems without reservation and taking comfort in the responses.

When dogs and other animals exhibit even small amounts of humanlike behavior, we tend to assume they are more like us than they really are. The same goes for machines, said Colin Allen, a professor at the University of Pittsburgh who explores cognitive skills in both animals and machines. “People get sucked in,” he said, “even if they know they are being sucked in.”

That is part of what is happening with GPT-3. Because it can generate convincing tweets, blog posts and computer code, we read humanity into this digital system — and pay less attention to its limits.

In practice, the system fails about as often as it succeeds. We overlook that the computer code it writes requires some fine-tuning from human programmers — a line removed here or added there. We do not notice that its talent for conversation breaks down after a few exchanges, when it cannot “remember” what it said just a few seconds before. We do not quite realize that although the system generated a convincing blog post for Mr. Porr, he provided the headline and the photo and the first few sentences, and he removed some sentences that were less convincing.

Mr. Porr does not believe GPT-3 is an enormous threat to the battle against disinformation in the short term, because it still requires so much help from humans. A tool like this becomes truly dangerous only if it can generate enormous amounts of convincing disinformation entirely on its own, exceeding what a team of hired hands can do with relative ease today.

Similarly, when app designers ask Mr. Singer of Square if GPT-3 is a threat to their careers, he assures them it is not — at least not yet. He sees it as a way of making their jobs easier. “If it can get 70 percent of the way there, that is a lot of tedious work taken out of the equation,” he said.

What we do not know is how much this technology will continue to improve in the months and years to come.

Smarter, faster, even more expensive
While the researchers at OpenAI were training GPT-3 on more than a trillion words posted to the internet, they ran a second experiment, training a similar system on tens of thousands of digital photos. That system could analyze all of those photos and learn to build images in much the same way that GPT-3 builds paragraphs. Given half of a cat photo, it could generate the rest of the cat.

For some researchers, the experiment indicates that such a system could ultimately handle tasks across multiple dimensions — language, sight, sound — much like humans do. Even when trained solely on language, they say, the system could already reach into other areas, whether computer programming, playing chess, or generating guitar tabs.

But continuing to improve this technology is far from trivial. Processing all of that internet data requires a specialized supercomputer running for months on end, an undertaking that is enormously expensive. When asked if such a project ran into the millions of dollars, Sam Altman, OpenAI’s chief executive, said the costs were actually “higher,” running into the tens of millions.

Mr. Amodei, OpenAI’s vice president for research, said there was still room to improve the technique, using more processing power to analyze more data. But he also said the approach might be close to running out of “juice.”

At the very least, GPT-3 is a new tool for a world of A.I. researchers and entrepreneurs, a way of building all sorts of new technologies and new products. Mr. Wrigley, the computer programmer, recently quit his day job to start a company called LearnFromAnyone, which aims to build a kind of automated tutor using GPT-3 that can assume the guise of everyone from scientist Douglas Hofstadter to venture capitalist Peter Thiel. Others are building companies that aim to automatically generate code for computer programmers and automatically write promotional emails and tweets for marketing professionals.

But it is unclear how effective these services will ultimately be. If GPT-3 generates the right text only half of the time, can it satisfy professionals? And it is unclear whether this technique is a path to truly conversational machines, let alone truly intelligent systems. Additional progress on the long road to machines that can mimic the human brain, Mr. Amodei said, will require entirely new ideas.

“It is kind of like a chemistry reaction,” he said. “We have this one ingredient. But other ingredients are required as well.”

(The New York Times)



UK Regulator Considers Opening Apple, Google App Stores to Rival Payments

FILE PHOTO: A Google logo is seen at a company research facility in Mountain View, California, US, May 13, 2025. REUTERS/Carlos Barria/File Photo
FILE PHOTO: A Google logo is seen at a company research facility in Mountain View, California, US, May 13, 2025. REUTERS/Carlos Barria/File Photo
TT

UK Regulator Considers Opening Apple, Google App Stores to Rival Payments

FILE PHOTO: A Google logo is seen at a company research facility in Mountain View, California, US, May 13, 2025. REUTERS/Carlos Barria/File Photo
FILE PHOTO: A Google logo is seen at a company research facility in Mountain View, California, US, May 13, 2025. REUTERS/Carlos Barria/File Photo

Britain's competition regulator on Tuesday proposed allowing app developers to steer users to alternative payment options outside Apple and Alphabet's Google app stores to cut fees and boost competition.

The Competition and Markets Authority said the proposals would remove restrictions that currently prevent UK developers from directing users to off-platform payment options, which are banned by Apple and restricted by Google.

The watchdog said any fees charged by two of the world's largest technology companies for allowing such "steering" would need to be fair and reasonable, and should be lower than current app store commissions, with savings passed on to consumers or reinvested in innovation.

"While it is only fair for Apple and Google ⁠to be compensated for ⁠the services they provide, any fees they charge must be justified through a robust, evidence-led framework involving due reference to both cost and value," Will Hayter, executive director for digital markets, is expected to say later on Tuesday, according to an excerpt of his speech.

The CMA said it was also considering requiring Apple to open up access to its near-field communication technology, which is used for contactless payments, potentially allowing developers to offer payment services within their own iOS ⁠apps.

This could enable UK fintech companies to build alternatives to Apple's wallet, including account-to-account payments and emerging technologies such as digital currencies, Reuters quoted the CMA as saying.

The proposals are part of a consultation under Britain's new digital markets regime, which gives the watchdog powers to impose tailored requirements on companies with so-called "strategic market status.”

Google said in an emailed statement it had already taken steps in that direction, pointing to new Play Store terms introduced earlier this month allowing developers to steer users to complete transactions outside the platform.

The CMA said it would assess Google's recent changes as part of its work before deciding later this year whether to impose formal requirements.

Apple has previously said it does not support allowing developers to direct users to off-platform payments, arguing this could undermine user ⁠security and fraud protections ⁠and limit its ability to verify transactions.

An Apple spokesperson said it could open the door to "scams, bait-and-switch tactics, and the circumvention of parental controls.”

"When users are directed away from Apple's trusted payment infrastructure, they lose the protections they rely on Apple to provide," the spokesperson said, adding the US tech giant would continue to "make our concerns clear" to the CMA.

The regulator designated Apple and Google as having strategic market status in mobile ecosystems last year, giving it the power to intervene more directly to boost competition.

In February, it secured commitments from the two companies to make their app stores fairer and more transparent, including changes to rankings, reviews and access to certain features – but they did not address commissions, which can reach up to 30%.

The CMA said at the time that enabling developers to steer users to alternative payment methods remained a priority, an issue that has also drawn scrutiny from regulators in the European Union, the United States and Japan.


Taiwan Raids Tech Firms in China AI Chip Smuggling Probe

This handout photo from the Taiwan Coast Guard taken on May 20, 2026 and released on May 21 shows pallets of servers made by Super Micro Computer seized by Taiwanese authorities at an undisclosed location. (Photo by Handout / TAIWAN COAST GUARD / AFP)
This handout photo from the Taiwan Coast Guard taken on May 20, 2026 and released on May 21 shows pallets of servers made by Super Micro Computer seized by Taiwanese authorities at an undisclosed location. (Photo by Handout / TAIWAN COAST GUARD / AFP)
TT

Taiwan Raids Tech Firms in China AI Chip Smuggling Probe

This handout photo from the Taiwan Coast Guard taken on May 20, 2026 and released on May 21 shows pallets of servers made by Super Micro Computer seized by Taiwanese authorities at an undisclosed location. (Photo by Handout / TAIWAN COAST GUARD / AFP)
This handout photo from the Taiwan Coast Guard taken on May 20, 2026 and released on May 21 shows pallets of servers made by Super Micro Computer seized by Taiwanese authorities at an undisclosed location. (Photo by Handout / TAIWAN COAST GUARD / AFP)

Taiwanese investigators have raided the Taiwan offices of US company Super Micro Computer and two other tech firms, a prosecutor said Tuesday, as part of an expanded probe into the alleged smuggling of Nvidia AI chips to China.

Prosecutors said in May they were investigating the shipment of "high-end" AI servers containing advanced Nvidia chips to China, Macau and Hong Kong, in violation of US export controls.

Nine people are now under investigation, up from three previously, Huang Sheng, head prosecutor in the Keelung Prosecutors Office, told AFP.

They are accused of forging documents so they could ship roughly 50 servers made by Super Micro Computer to China.

Some of the servers were cleared by Taiwan customs and sent to China via Japan, an official previously told AFP on the condition of anonymity.

Twelve sites were raided on Monday as part of the probe, the prosecutors office said in a statement.

They included the homes of six people and offices of the companies they worked for -- Nasdaq-listed Super Micro Computer and Taiwan-listed firms Albatron Technology and Chief Telecom.

The United States restricts the export of its most cutting-edge AI chips to China, partly over concerns the technology could be used by Beijing's military.

But it is not a criminal offence in Taiwan -- a situation lawmakers and experts say needs to change -- with Taiwanese prosecutors relying on other laws to go after offenders.

Lawmaker Chung Chia-pin, who belongs to President Lai Ching-te's Democratic Progressive Party (DPP), plans to propose an amendment to the Foreign Trade Act to include a "mainland China semiconductor chip clause" that would make exporting chips there illegal.

Chung told AFP Tuesday that a loophole in the law was created under former president Ma Ying-jeou, who belongs to the Kuomintang party, and successive DPP-led governments have failed to close it.

Top-end chips made by US titan Nvidia -- the world's most valuable company -- are used to train and run AI systems.

In response to Washington's export restrictions, China has been accelerating efforts to develop its own AI chips and break away from reliance on US hardware.

This month, Taiwanese Deputy Economic Affairs Minister Ho Chin-tsang said Taiwan and the United States "will work to implement our shared export control goals", but the government has not provided details.

Chris McGuire, an expert on China and AI at the US-based Council on Foreign Relations, said chip smuggling was a "really significant problem" in Taiwan and Southeast Asia.

"It's really, really important that allies align with the United States on all of these policies and also legal authorities," McGuire, who worked at the National Security Council under former US president Joe Biden, told a forum in Taipei this month.

"It's not a criminal violation in Taiwan to export AI chips to China, obviously it is under US law, but it's not under Taiwanese law. That needs to change, right?"

Super Micro Computer, Albatron Technology and Chief Telecom have said separately they are cooperating with investigators. Their shares have seen sharp falls this week.

Prosecutors say it is too early to know if the case is linked to a Nvidia chip smuggling case involving Super Micro Computer employees in the United States.

A US indictment unsealed in March showed employees of the company allegedly raked in billions of dollars diverting Nvidia AI chips to China in breach of export controls.


WhatsApp Will Allow Users to Go by Usernames Instead of Phone Numbers, Closing a Privacy Blind Spot

A WhatsApp icon is displayed on an iPhone, Nov. 15, 2018, in Gelsenkirchen, Germany. (AP)
A WhatsApp icon is displayed on an iPhone, Nov. 15, 2018, in Gelsenkirchen, Germany. (AP)
TT

WhatsApp Will Allow Users to Go by Usernames Instead of Phone Numbers, Closing a Privacy Blind Spot

A WhatsApp icon is displayed on an iPhone, Nov. 15, 2018, in Gelsenkirchen, Germany. (AP)
A WhatsApp icon is displayed on an iPhone, Nov. 15, 2018, in Gelsenkirchen, Germany. (AP)

WhatsApp users will soon get the option of going by usernames instead of phone numbers, the company said Monday, announcing plans to address a privacy blind spot.

The app said it has started allowing users to reserve unique usernames, which can be used to contact WhatsApp users when the feature is launched later this year.

WhatsApp, which says it has more than 3 billion users globally, has until now allowed users to be contacted by anyone who has their phone number.

The app, owned by Meta Platforms, said in a blog post that over the “coming months” users will get the option to be found and contacted only by their username, and not their number. It wasn't more specific about the timeline.

“We have designed this as a core privacy feature,” Alice Newton-Rex, WhatsApp's vice president of product, told reporters.

There won't be a directory of usernames on the app, and the app won't suggest names as you type.

“People will need to know your exact username to contact you for the first time,” she said.

WhatsApp's current privacy settings are limited to blocking individual users and silencing unknown callers. The app also allows users to add a profile name, but that's only displayed in chat groups for other people who don't have the user's contact info saved.

While Americans still prefer text messaging to WhatsApp, the app is widely used in Europe, Asia and much of the rest of the world.

Catchy online handles are highly coveted and users will likely scramble to claim a desirable one.

“I think a lot of people will go and get usernames and that’s why we decided to open reservations early,” Newton-Rex said.

Companies, organizations and creators with existing accounts on Meta's social media platforms, Instagram and Facebook, will get the chance to claim their usernames on WhatsApp.

Usernames need to be between three and 35 characters. To prevent impersonation, WhatsApp will hold back usernames for high-profile people or groups such as celebrities, public figures and government entities.