By David Ingram and Kevin Collier -
NBC News
OpenAI, the San Francisco, California-based tech company that grabbed global attention when it launched ChatGPT, announced on Tuesday that it was unveiling a new version of its Artificial Intelligence (AI)
software
.
Called GPT-4, the tool "can solve difficult problems more accurately, thanks to its broader general knowledge and problem-solving capabilities," OpenAI explained in an announcement on its website.
[Microsoft's search engine with artificial intelligence can insult its users, threaten them and compare them to Hitler]
In a demo video, Greg Brockman, president of OpenAI, showed how the technology could be trained to quickly answer tax-related questions, such as calculating the standard deduction and total tax liability for a marriage.
“This model is very good at mental math.
It has extensive capabilities that are very flexible,” he assured.
And in a separate video the company posted online, it explained that GPT-4 had a number of capabilities that the earlier version of the technology didn't, including the ability to "reason" based on images users have uploaded. .
“GPT-4 is a large multi-modal model (accepts images and text, and outputs text) that, while less capable than humans in many real-world scenarios, shows human-level performance in various professional benchmark tests and academics”, published OpenAI on its website.
[Four astronauts land off the Florida coast at the end of a five-month mission]
Andrej Karpathy, an OpenAI employee, tweeted that the feature meant the AI could "see."
New technology is not free, at least for now.
OpenAI says you can try GPT-4 on its subscription service, ChatGPT Plus, which costs $20 a month.
OpenAI and its ChatGPT chatbot have shaken up the tech world and alerted many outsiders to the possibilities of AI software, thanks in part to the company's partnership with Microsoft and its search engine, Bing.
But the pace of OpenAI releases has also caused concern, because the technology is unproven and forces abrupt changes in fields ranging from education to the arts.
The rapid public development of ChatGPT and other generative AI programs has led some ethicists and industry leaders to call for limits to this technology.
Sam Altman, chief executive of OpenAI, tweeted Monday that "we definitely need more regulation on AI."
The company detailed GPT-4's capabilities in a series of examples on its website: the ability to solve problems, such as scheduling a meeting between three busy people;
earn high scores on tests, such as the uniform bar exam;
and learn a user's creative writing style.
ChatGPT, an impressive technological tool that will change the lives of many people
Feb 6, 202302:04
But the company also acknowledged limitations, such as social biases and "hallucinations" that it knows more than it really does.
Google, concerned that AI technology could cut into the market share of its search engine and cloud computing service, launched its own
software
, known as Bard, in February.
OpenAI was launched in late 2015 with the backing of Elon Musk, Peter Thiel, Reid Hoffman and tech billionaires, and its name reflected its status as a non-profit project that would follow the principles of open source software shared freely
online
. .
In 2019, it moved to a lucrative “limited” model.
[What would you do if your date refused to pay $3 extra cheese on the burger? A girl goes viral on TikTok with this false story]
Now it launches GPT-4 with some secrecy.
In a 98-page document accompanying the announcement, company employees say they will keep many details under wraps.
In particular, the data on which the model is based will not be made public.
“Given the competitive landscape and security implications of large-scale models such as GPT-4, this report does not contain further details on the architecture (including model size), hardware, training computation, construction of the data set, training method, or the like,” they wrote.
How does the use of filters in photos affect self-esteem and self-perception?
This is what an expert thinks
March 9, 202304:24
They added: “We plan to make more technical details available to additional third parties who can advise us on how to balance the aforementioned security and competitive considerations against the scientific value of increased transparency.”
The release of GPT-4, the fourth iteration of OpenAI's foundational system, has been rumored for months amid growing anticipation around the chatbot that builds on it.
In January, Altman lowered expectations for what GPT-4 could do, telling the
StrictlyVC
podcast that "people are begging to be let down, and they will."
[USA. aims to make chips to limit dependence on Taiwan. Why can cell phone prices go up?]
On Tuesday he asked for opinions.
“We've had the initial lineup of GPT-4 done for a while now, but it's taken us a long time and a lot of work to feel ready to launch it,” Altman posted on Twitter.
“We hope you enjoy it and we really appreciate feedback on its flaws,” he added.
Sarah Myers West, managing director of the AI Now Institute, a nonprofit group that studies the effects of AI on society, said releasing such systems to the public without oversight "is essentially experimenting in the wild."
“We have clear evidence that generative AI systems routinely produce error-prone, derogatory, and discriminatory results,” he said in a text message.
“We can't just rely on companies' claims that they will find technical solutions to these complex problems,” he said.
