Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
Tim Guleri, a venture capitalist at Silicon Valley firm Sierra Ventures, has witnessed numerous technology revolutions throughout his career, but recently told VentureBeat he believes the current emergence of generative AI is the most significant and intimidating of them all.
With three decades of experience in founding and investing in enterprise technology, Guleri built companies during the Internet boom of the 1990s and then invested in startups during the subsequent Web 2.0, mobile and big data revolutions of the 2000s and 2010s.
Guleri’s investments include Sourcefire and MakeMyTrip, which eventually went public. In 2000, he founded Octane Software, an e-commerce company that sold for $3 billion to Epiphany while he was serving as CEO.
Guleri’s resume establishes him as a VC heavyweight in the area of enterprise software investing.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
Discussing the current investment landscape, Guleri told VentureBeat in an exclusive interview: “I have never been as excited and as scared in my 20 years of doing venture capital because of gen AI.”
His excitement stems from the potential for generative AI to vastly improve value creation in the industry, while his fear arises from concerns about privacy and ethical issues surrounding the technology.
“I’m excited because everything that we have done to create value in the industry can be redone 100x better riding on the back of gen AI,” he continued. “I’m scared because of the things that we’re talking about in terms of hallucinations and privacy and all of the things that I believe need to be fixed entirely… and I’m a bit paranoid, I guess, when it comes to the firm because I think we need to lead the way as an early stage venture firm in the way we operate — to catch the best opportunities early enough.”
His marketing partner at Sierra, Anne Gherin, said Guleri has for the past few months spent nights and weekends learning, testing and investing in generative AI. Every Monday partner meeting, he brings in a new report, presentation or diagram on the topic, she said.
As the recent breakthroughs of ChatGPT have garnered widespread attention, Silicon Valley has increasingly embraced AI, with many heralding generative AI as an unparalleled paradigm shift. Marc Andreessen, co-inventor of the first web browser and leader of prominent venture firm Andreessen Horowitz (a16z), recently published a blog post asserting that AI could be the most important and beneficial invention in human history, on par with or surpassing electricity and microchips.
However, big firms like Andreessen’s can make big shifts based on trends. Until last year, it dominated investments in crypto and other financial services.
Meanwhile, Guleri’s focus on early-stage investing in enterprise technology hasn’t wavered, through both frothy and bear markets.
While his approach can yield substantial returns if he invests in the right companies, it also carries significant risk due to the challenge of identifying promising startups.
That’s what underscores his obsession with generative AI.
Gen AI is changing the company-building playbook
Guleri says generative AI will disrupt the traditional playbook of building companies in the enterprise software market for the past 30 years, one that has centered around automation. Successful companies like Oracle, PeopleSoft, Workday, Salesforce.com, and ServiceNow have all been built by taking relatively manual processes and automating them.
Whether it was customer support, sales, IT, or human resources, the approach started with a relationship database, along with a business process on top of it. Guleri explained: “It’s a UX slapped on a database table,” and a bunch of sales people that know how to sell.
Guleri believes generative AI will pull apart that software ecosystem and reassemble it, and do so in unpredictable ways, he said.
Like some industry analysts, he uses the word “intelligence” to describe how generative AI will elevate technology, referring to the ability of large language models (LLMs) to mimic aspects of human intelligence.
“The next multiple decades will be about intelligence,” he said, not automation.
The next enterprise software blueprint is emerging, and includes more open source
Guleri emphasized the necessity for his firm to maintain an open mind while formulating a vision for the future of enterprise architecture.
Guleri shared his contrarian view that business applications will continue to be built on top of relational databases, rather than vector databases, which have been touted for their ability to handle unstructured data that drives LLM models. He believes relational databases will integrate vector and embedding capabilities, eliminating the need for separate vector databases.
Another clear pattern is that there will be no LLM to rule them all. “There’s this notion of chaining LLMs,” he said.
Moreover, Guleri predicted that open source will dominate large language models (LLMs) due to increasing regulatory scrutiny on companies like Amazon, Google, and OpenAI. Those innovators will have to be more “premeditated” in what they put on the market and enterprise customers will be wary about what they put into production. “Open source is going to run ahead,” he said.
He also highlighted the importance of identifying genuine generative AI startups amid a sea of pretenders. Many entrepreneurs have “ben AI-washed” their presentation decks from six months ago, he said.
With about 13 investments in generative AI under its belt, Sierra Ventures began backing such startups before the recent surge in interest. The firm’s first investment in the field came in 2018 with Krisp, a noise-canceling app startup, followed by a $4.25 million seed round in Quillbot, a content management company, in 2020. Quillbot was later acquired by Course Hero for an undisclosed sum. Other investments include Deephow and Modulate.
Guleri acknowledged that the disruptive power of generative AI has only recently become apparent, stating, “It has only become obvious to me over the last six months – on how fundamentally disruptive this is going to be.”
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.