Venture capitalists are racing to invest in artificial intelligence start-ups as growing hype around “generative AI” fills the void left by failed cryptocurrency and blockchain companies.
The recent leap in the development of sophisticated computer programs capable of writing scripts and creating art in seconds has sparked renewed investor interest, creating a rare bright spot in a start-up landscape dominated by plummeting valuations and job cuts.
OpenAI, a San Francisco-based company in which Microsoft is the largest backer, last week released the latest version of its GPT-3.5 software, which can chat with users via text: answer follow-up questions, admit mistakes and reject inappropriate requests.
In five days, ChatGPT surpassed one million users and was praised by billionaire Elon Musk, co-founder of OpenAI who left the board in 2018, who tweeted: “ChatGPT is scary. We are not far from a dangerously powerful AI.
The freedom to tinker with such powerful AI has already sparked new start-up ideas for countless investors and entrepreneurs.
“Any organization can pick them up and start training them and playing with them, see what they can produce,” said Ed Stacey, managing partner at IQ Capital. “It really doesn’t make sense to sit on the sidelines anymore.”
As Silicon Valley investors’ social media feeds fill with examples of AI-generated images and text, pivot to AI comes as interest in so-called “Web3” , a vision of decentralized virtual worlds built on blockchains, is shrinking amid the recent cryptocurrency crash.
“There’s a huge hype cycle going on,” said Colin Treseler, co-founder of Supernormal, which uses AI to summarize online meetings. “The Web3 hype died down and these people needed a place to go.”
Venture capital investment in generative AI has grown 425% since 2020 to $2.1 billion this year, according to data from PitchBook, even as the broader tech market declines.
An artificial intelligence entrepreneur says after discussing a possible fundraising with just three investors, he was inundated with offers from more than 20 others, securing funding after a week of flash meetings around the world .
Two deals in mid-October marked the start of the latest funding frenzy, which saw investments made by Coatue Management and Lightspeed Venture Partners.
Jasper, who describes himself as an “AI copywriter” to marketers, raised $125 million at a $1.5 billion valuation, while one of the companies behind the tool image generation Stable Diffusion, London-based Stability AI, has raised $101 million, in a move that has seen it achieve “unicorn” status – a $1 billion valuation. This week, another Stable Diffusion developer, Runway, raised $50 million.
When Cristóbal Valenzuela co-founded Runway four years ago, investors told him, “Generative AI is not a thing.” “Everyone thought we were a little crazy,” he said. Now investors say the technology could be “as transformative as mobile was 20 years ago,” he added.
“What has changed more recently is that the quality of the models has gotten really good,” Valenzuela said. “It’s no longer about [saying], ‘imagine a future where this could happen’. It’s happening now.
ChatGPT and OpenAI’s DALL-E art program, alongside rival graphics tools Stable Diffusion and Midjourney, are examples of using large language and image models, sometimes called generative AI or base models, which can produce content based on previous word sequences or images.
In September, Sequoia Capital partners co-wrote an investment thesis using GPT-3 software, claiming that AI would be able to produce better-than-human-average final writing drafts, generate commercial-scale code and draft to images and games within the next couple of years.
“Generative AI is set to become not only faster and cheaper, but in some cases better than what humans create by hand,” Sequoia’s analysis concluded.
The company invested in Hugging Face in May as part of a Series C funding round that valued the startup, which has its own big language model Bloom, at $2 billion.
The global market for AI-augmented content solutions is expected to reach $2.3 billion this year and is expected to grow globally by 17% through 2025, according to research by PitchBook. But the group added that “the technology may not generate high revenues in the short term as professions resist AI solutions and the technology has yet to mature.”
Due to the high costs of running programs and storing large amounts of data that AI programs learn from, large language models were previously reserved for tech giants like Microsoft, Google, and Facebook.
But OpenAI has made its technology available through an application programming interface, giving any business access to its capabilities.
Founded in 2015, OpenAI was created as a non-profit organization founded on the principles of making AI accessible to everyone and developing the technology safely – the brainchild of some of the world’s most radical thinkers. technology, including Musk and Peter Thiel.
In 2019 it became a for-profit company, which was shortly followed by a billion-dollar deal from Microsoft, including the use of its Azure cloud computing platform to conduct experiments. As part of the deal, Microsoft gets a first chance to commercialize early research results from OpenAI.
Microsoft’s focus on OpenAI was part of a drive to reclaim an edge in AI from rival Google’s massive investments in using search and speech technology, as well as in the acquisition of London-based AI firm DeepMind for around £400m in 2014.
“One of the big historical advantages for companies is having access to large datasets that are usually proprietary. They have been able to use those datasets to train bigger and bigger models,” said Stacey of IQ Capital.
On average, the cost of running ChatGPT is estimated to be pennies per chat, according to OpenAI chief executive Sam Altman. Asked on Twitter if the tool would be free forever, he replied, “We’ll have to monetize it somehow at some point; the computational costs are exorbitant.
we will have to monetize it somehow at some point; computational costs are exorbitant
—Sam Altman (@sama) December 5, 2022
It has proven so popular in recent days that the platform has limited who can use it. The company is raising more money from investors, according to The Information, and Altman tweeted this week that it was looking to hire more staff.
Bloc Ventures, a UK-based high-tech venture capital firm, has focused its investments on technologies that enable high levels of cloud computing, reducing both cost and energy used in generative AI .
“We are in a world where companies are looking for net zero [carbon emissions]and the luxury of having chatbots that we can talk to via AI is digging a hole in the ground in a data center,” said David Leftley of Bloc Ventures.
New York-based Runway is taking a more ambitious and expensive approach, both doing the primary AI research to build models and turning it into a suite of imaging and collaboration tools already used by companies such as Publicis, Google and CBS.
“Many companies rely on existing APIs [but] our long-term bet is that you have to own your stack, you have to own your technology, to enable you to switch faster and easier when needed,” Valenzuela said.
He insisted that AI start-ups like Runway could outsmart big tech companies: “The field is moving extremely fast. The pace of learning really matters – how you adapt and how you change.
By making ChatGPT open source and publicly available, OpenAI is able to collect more data to train its large language models and fix bugs.
A significant limitation of the technology, and similar AI tools, are so-called “hallucinations,” where the program gives a convincing answer that is factually inaccurate and struggles with simple math. An example posted on Twitter said ChatGPT incorrectly claimed that Angela Merkel and Gerhard Schröder belong to the same political party.
“The potential for misinformation to spread is huge,” said Carissa Véliz, associate professor at the University of Oxford’s Institute of Ethics and AI. “If you ask him to create a Covid conspiracy theory, that can make a really compelling case.”
OpenAI admits that GPT “sometimes writes plausible but incorrect or nonsensical answers,” among other limitations of the technology, leading many to suggest that the technology requires human intervention before it can be incorporated into businesses.
“There are a lot of questions about the commercial viability of these models and capabilities,” said Lisa Weaver-Lambert, head of private equity, data and AI at Microsoft, who said generative AI was in an “experimental phase”.
She added, “If I was looking to invest in this space, I would think about what are the concrete business problems that actually exist that people have solved. [and can AI] find a faster and cheaper way to accomplish the same thing? »
Additional reporting by Madhumita Murgia in London
#Investors #seek #cash #gamechanging #generative #startups