Not everyone is convinced of the return on investment of generative AI. But many investors are, judging by the latest figures from funding tracker PitchBook.
In the third quarter of 2024, venture capitalists invested $3.9 billion in generative AI startups across 206 deals, according to PitchBook. (That's not counting OpenAI's $6.6 billion round.) And $2.9 billion of that funding went to U.S.-based companies in 127 deals.
Some of the biggest winners in the third quarter were coding assistant Magic ($320 million in August), enterprise search provider Glean ($260 million in September), and business analytics company Hebbia ($130 million). dollars in July). China's Moonshot AI raised $300 million in August, and Sakana AI, a Japanese startup focused on scientific discovery, closed a $214 million tranche last month.
Generative AI, a wide range of technologies that includes text and image generators, coding assistants, cybersecurity automation tools, and more, has its detractors. Experts question the trustworthiness of the technology and, in the case of generative AI models trained without permission on copyrighted data, its legality.
But venture capitalists are effectively betting that generative AI will take hold in large, profitable industries and that its long-term growth will not be affected by the challenges it faces today.
Maybe they are right. A Forrester report predicts 60% of generative AI skeptics will adopt the technology, knowingly or not, for tasks ranging from summarization to creative problem solving. That's considerably more optimistic than Gartner's. prediction At the beginning of the year, 30% of generative AI projects will be abandoned after proof of concept by 2026.
“Large customers are deploying production systems that leverage startup tools and open source models,” Brendan Burke, senior emerging technology analyst at PitchBook, told TechCrunch in an interview. “The latest wave of models shows that new generations of models are possible and can excel in scientific fields, data retrieval and code execution.”
A formidable obstacle to widespread adoption of generative AI is the technology's enormous computational requirements. Bain analysts project in a recent study that generative AI will drive companies to build gigawatt-scale data centers (data centers that consume 5 to 20 times the amount of energy that the average data center in the world consumes). current), which will strain an already strained electricity and labor supply chain.
AI-driven data center power generative demand is already increasing. prolonging the life of coal-fired plants. Morgan Stanley estimates that if this trend continues, global greenhouse gas emissions by 2030 could be three times higher than if generative AI had not been developed.
Several of the world's largest data center operators, including Microsoft, Amazon, Google and Oracle, have announced investments in nuclear energy to offset their growing consumption of non-renewable energy. (In September, Microsoft said it would harness power from the infamous Three Mile Island nuclear plant.) But it could take years before those investments pay off.
Investments in generative AI startups show no signs of slowing down – negative externalities be damned. ElevenLabs, the viral voice cloning tool, is reportedly looking to raise funds at a valuation of $3 billion, while Black Forest Labs, the company behind the notorious X image generator, is said to be in talks for a financing round of 100 million dollars.