Everyone was talking about EVs for a while, and no one was paying attention to data centers. Now data centers have come out of nowhere.
Some analysts predict a VC comeback in 2025, while others believe global economic uncertainty will keep investments paused.
We spoke with Dr. Rakesh Radhakrishnan, operating partner of Prithvi Ventures—investors in early-stage, hyper-ambitious startups fighting climate change—about the outlook for VC, exits, and investing trends in 2025.
Exit mindset: Radhakrishnan approaches potential acquisitions with a growth mindset rather than prioritizing exits. “You can’t go in and say, ‘I’m gonna get this exit in six years,’” he explains. “Your first objective should be: How am I going to grow this business in six years? How am I going to create a product? How am I going to create revenues? How am I going to become profitable?”
Once that plan is in place and a team demonstrates execution capability, the “exit will take care of itself,” Radhakrishnan says.
Data centers: Strain on the electric grid is far from a new concern. The infrastructural strain from data centers has risen significantly in the past 12-24 months as LLMs have gone fully mainstream. Pointing to AI’s effects on climate-related investments, Radhakrishnan notes, “Five years ago, everyone was talking about EVs and their strain on the electric grids, and no one was paying attention to data centers. Now data centers have come out of nowhere. There are many cash-rich companies actively investing huge dollars in that marketplace. There’s a need for solutions that can provide cleaner power to data centers and help with more efficient cooling, more effective communications, and the entire power infrastructure of data centers.”
There’s a need for solutions that can provide cleaner power to data centers and help with more efficient cooling, more effective communications, and the entire power infrastructure of data centers.
M&A outlook: “The thresholds for what an emerging company has to hit for larger companies to buy them have not changed in these twenty-plus years,” Radhakrishnan explains. “M&A may have slowed down, perhaps because of regulatory challenges, but those companies are still sitting on cash and are willing to deploy if the risks have been appropriately reduced to a level where they feel they can take this on and scale the business.”
AI and ML model investments: Radhakrishnan asserts that while investment in AI and ML models is substantial, adoption in energy and climate tech lags. “You have to show that this particular toolset you’re developing is better than the software solutions that current engineers and scientists in those lab and engineering facilities are using today,” he explains. “Those solutions have a legacy of twenty-plus years and strong stickiness in terms of what has been developed. Displacing that and showing that you have something better is going to create an extended timeline.”
Quantum computing: “Quantum computing has gotten some attention recently,” Radhakrishnan adds. “There’s a promise of quantum being able to solve problems that even those advanced AI and ML tools may not reach. The challenge is that the hardware is further along in terms of investments. Practical solutions require not just algorithms and a software layer, but actually tying them to real applications in industry where they could benefit. This is under the radar right now, but you’re going to hear more over the next year or so.”
