Samuel Altman, CEO of OpenAI, testifies before the Senate Judiciary Subcommittee on Privacy, Technology, and the Law May 16, 2023 in Washington, DC
Explanations, please: Sam Altman, chief executive of San Francisco-based start-up OpenAI © Win McNamee/Getty Images

Everyone seems to agree that generative artificial intelligence is the shiny new thing in technology and will transform the world. The trouble is that no one can yet explain exactly how.

The US venture capital firm Sequoia suggests the technology is evolving fast and going through its “awkward teenage years”. But some of the most powerful companies in history, including Google, Microsoft, Meta and Amazon, as well as scores of start-ups, are investing billions into figuring out what generative AI will do when it grows up.

San Francisco-based start-up OpenAI triggered the latest investment frenzy in November 2022 with its launch of the ChatGPT chatbot. Although tech companies had been playing around with generative AI services for years, ChatGPT’s arrival ensured the technology broke into the mainstream.

It seemed that every chief executive of every company — as well as every student trying to turn in a last-minute essay — was wowed by ChatGPT’s remarkable ability to generate reams of plausible text at lightning speed, promising to unleash creativity and boost productivity. OpenAI said 100mn users had experimented with ChatGPT within the first two months of its launch — one of the fastest take-ups of any new product in corporate history.

Investment has since poured into the sector with AI companies pioneering not just text but image, audio and video generation services, too. CB Insights, the New York-based data firm, estimated $17.4bn had been invested into 170 generative AI companies in the first nine months of 2023 — including Microsoft’s $10bn investment into OpenAI.

Several other big tech companies have also sought tie-ups with promising generative AI start-ups: Google and Amazon have both invested in Anthropic, while Microsoft and Nvidia have backed Inflection AI.

Established companies across many industries, from healthcare to financial services to defence, have been keen to deploy the technology. By the end of the third quarter of 2023, generative AI had been mentioned on 2,081 corporate earnings calls, compared with none over the same period the previous year, according to CB Insights.

Many tech analysts reckon that generative AI represents the next big platform shift in computing, following the mobile and cloud computing revolutions. It will eventually provide the new interface between humans and machines as we all adopt AI-enabled digital agents to manage our lives.

The futuristic world depicted in the 2013 science fiction film Her, in which Theodore Twombly increasingly relies upon — and eventually falls in love with — his virtual assistant Samantha, is now hoving into view.

A  computer screen with the home page of the artificial intelligence OpenAI web site, displaying its chatGPT robot
The launch of ChatGPT in November 2022 triggered an investment frenzy © Marco Bertorello/AFP via Getty Images

The infrastructure layer of generative AI is likely to be dominated by the leading AI companies, such as OpenAI and Google, that can afford to invest capital at massive scale to train, and run, the large language models that power the chatbots. But the application layer is likely to be far more competitive, as many other companies develop smaller, cheaper, more specialised and, often, open-source, models, focusing on a particular domain.

Silvio Savarese, Salesforce chief scientist and Stanford University computer science professor, says these smaller models will become more “agentic”, enabling them to be more proactive: “We believe that smaller models are more effective than large monolithic models, such as ChatGPT. We can focus on much more specific questions and topics.” He adds: “The idea is that these specialist models are far easier to train, have far fewer parameters, and can be deployed on mobile devices.”

Apple, the world’s most valuable company, which had been lagging behind some Silicon Valley rivals in generative AI, is now developing plans to deploy the technology on mobile devices. Such a transition would open up a new realm of applications as users could access generative AI on the go.

But generative AI’s rapid emergence has aroused governance concerns. November’s boardroom row at OpenAI, wherein chief executive Sam Altman fired and rehired within a week, has unsettled many in the industry.

Can such companies be trusted to develop such a powerful general purpose technology on their own? Generative AI systems raise a host of concerns about job displacement, copyright infringement, and the profusion of disinformation, plus the erosion of privacy and the concentration of corporate power. Regulators around the world have been rushing to catch up with the latest technological breakthroughs.

In November, the US administration signed an executive order instructing almost every federal agency to modify their rules to take account of generative AI. The UK government hosted an international summit at Bletchley Park to discuss the existential risks of AI. And, in December, the EU passed an act banning AI’s use in some “high-risk” areas and regulating it in many others.

Unless the generative AI industry grows up fast, governments are likely to set them even stiffer exams in future.

Copyright The Financial Times Limited 2024. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article