No innovation has had a greater cultural impact and no technology product has made a bigger splash in the past six months than OpenAI‘s ChatGPT.
The Microsoft-backed startup’s generative artificial intelligence chatbot wowed consumers when it debuted at the end of November. It revealed a quantum leap in the ability of humans to seamlessly interact with AI, which in turn can access the entire information universe of the internet.
AI is having its iPhone moment. Apple’s breakthrough product sparked the invention of a new ecosystem of apps bringing users new services, ranging from Uber to Instagram, because suddenly they had a computer in their pocket. Coinciding with that mobile revolution was a computing one as well, with exponential power to shift data into the cloud.
Now we’re seeing a similar technological boom around AI. It’s not just about the startling experience of interacting with the latest chatbots. AI will influence, disrupt and accelerate every industry. In fact, it’s already happening.
With OpenAI topping this year’s Disruptor 50 list, there’s no question that the dominant theme not just for the annual ranking but for the venture-backed tech startup space as a whole is artificial intelligence.
And it’s not just companies that have AI at their core. We’re seeing a range of enterprise applications for AI to drive efficiency and new capabilities across companies and sectors of the market. Of the 50 companies on this year’s list, 21 told us that AI is critically important to more than 50% of their revenue.
Half of the companies in the top 10 of the 2023 CNBC Disruptor 50 list feature key use of AI, and notably, they represent a diverse range of industries and use cases. Canva, the No. 3 company, is integrating ChatGPT into its design tools, giving customers a new way to be creative. No. 4 Disruptor Relativity Space is using AI to make 3D-printed rockets. No. 7 Disruptor Anduril Industries deploys AI to identify and attack security threats. U.K.-based renewable energy company Octopus Energy, No. 8 on this year’s list, uses AI to efficiently match energy supply and demand. No. 9 Lineage Logistics uses AI to optimize the movement of goods across the temperature-controlled supply chain.
“I do think we are deep into a new technological wave and this is, I think, the biggest one in a while,” OpenAI CEO Sam Altman said in an interview with CNBC late last week.
No. 19 on the Disruptor 50 list, Scale AI, has worked with companies including OpenAI to label the massive amounts of data — images, text, voice and video — that the machines need to digest to become better learners. Also on the list is the No. 44 Disruptor, Cohere, which was founded by former Google Brain researchers who helped develop a new method of natural language processing — transformers — that enable systems to grasp a word’s context more accurately.
Altman said OpenAI is seeing artificial intelligence affect nearly every industry. He pointed to the legal profession as a prime example.
“What we’re hearing from customers using our API for legal companies is that it is totally transforming the way they work and the efficiency that any one lawyer can achieve and the accuracy, freeing people up to do more of what they do really well, and having this new tool to sort of give them as much leverage as possible,” Altman said.
“That is a pattern we’re seeing again and again in many industries, and I’m super excited about it,” he said.
Its ability to make stock market investors skittish became clear when Alphabet‘s shares tanked after the rollout — which some employees called rushed — of its ChatGPT competitor, Bard, earlier this year. And in one of the sectors seen as being most acutely at risk from generative AI, education, Chegg saw its shares fall by close to half just because its CEO referenced an impact from ChatGPT on customer growth during its recent earnings call.
For now, OpenAI has a dual revenue stream: an enterprise software model where it charges companies for access to the platform, and a premium chat app it offers to consumers for $20 monthly, in addition to the free version.
“For now, we’re pretty happy with these two models. We’re super open to explain other things,” Altman said, “you know, when we’re very much at the very start of this technology.”
OpenAI’s business customers include Salesforce, Snapchat, and its backer Microsoft, which is bringing OpenAI’s generative AI technologies to its Bing and Edge internet browsers and Microsoft 365 suite of business software, including Word, PowerPoint and Excel.
Microsoft’s cumulative investment in OpenAI has reportedly swelled to $13 billion, and the startup’s valuation is reported to be as high as $29 billion. The company declined to provide any funding or valuation data.
The growth in the power of AI has been so rapid and dramatic it has sparked concern from politicians and regulators. Those looking to play in the space — including Elon Musk, who was an early co-founder of OpenAI and now says he will launch a competitor — are also speaking out about the risks. Musk, along with Apple co-founder Steve Wozniak and a range of professors and CEOs, signed an open letter in March from the Future of Life Institute, urging AI labs to stop training models that are more powerful than OpenAI’s GPT-4.
Altman first responded in an appearance at a virtual event at MIT, saying that consistent safety guidelines were needed but that this proposed pause was “missing most technical nuance about where we need the pause.”
Altman continues to advocate for regulation. “We really need regulation here. We’ve been calling for it since the start of the company,” he said. “I think we’re going to get some regulation, and we’ll get more over time. And I think that’s really critically important. So I’m happy that it’s happening.”
“I think to get to the future where we have as much of the good use of AI and minimize the what could be quite bad uses of AI,” Altman said, “there’s just no way around having regulation here. We have regulation for other industries with much less powerful technology. So we should definitely have it here.”
Reid Hoffman, partner at venture capital firm Greylock, was an early investor in OpenAI and is now an investor in a number of AI companies and the co-founder of AI startup Inflection. He said he finds some of the criticism to be more dangerous than OpenAI.
“A bunch of it is well-intentioned; there are a bunch of different ways AI can play out,” said Hoffman, who is also on the Microsoft board of directors and had served on the OpenAI board before stepping down due to potential conflicts of interest. “Some of it is less well intentioned: ‘Everyone else, slow down so I can speed up.’ And this is one of those things where it is overall a mistaken effort. … The call to slow down is, in fact, less safe than what they’re proposing,” he said, referring to OpenAI and Altman.
In addition to concerns about AI being used to manipulate or mislead, Altman said he is working to tamp down on bias within OpenAI’s systems.
“A big part of that is what we call RLHF, or reinforcement learning from human feedback, where we take these models that are pretrained on a significant fraction of the internet and we can sort of push them in certain ways,” Altman said. “We can teach the models like, ‘Hey, there’s a bias here in the data. You shouldn’t act this way.'” He said that from GPT-3 to GPT-4 the company has been able to make great strides in reducing bias in the model.
As companies including OpenAI battle bias and push for smart regulation, they’re also working with the established tech behemoths, such as Microsoft, and leaders in all sorts of industries to help them evolve, so they’re not disrupted.