The conversations around AI and sustainability have centered around two fronts, according to Lorenzo Thione, a managing director at Gaingels.

On the one hand, AI’s capabilities have grown rapidly, giving rise to new products and technologies that require enormous amounts of computational power and resources. That’s not quite sustainable.

But on the other hand, these advancements in AI and machine learning tech have widened the avenue for creating more architectures to sustain AI development. That, per Thione, is going to herald innovation.

Thione, like everyone these days, has lots of things to say about AI. However, he’s been in the game much longer than most people, having sold his AI-web company, Powerset, to Microsoft in 2009. Then he became an investor and has since become known for his work with Gaingels, a venture capital syndicate where he leads the AI and sustainability investments.

The syndicate, which has nearly $750 million deployed, focuses on backing founders who identify as part of the LGBTQ+ community. As a gay man, Thione remembers a time when founders used to be nervous about being open and out with investors, or worried about how they presented themselves. “They simply were not getting any access to the venture engine because of who they were,” he told TechCrunch+.

Thione and others have worked hard to change that narrative, and the venture industry is finally showing signs of entering a new era. Earlier this month, Crunchbase said it would begin tracking the number of venture dollars allocated to LGBTQ+ founders, which is estimated to stand at 1%.

Thione hailed Crunchbase’s move as a step towards tangible equity. “We cannot begin to fix what we do not measure,” he told TechCrunch+. “This could be the catalyst we need, fostering more diverse and successful LGBTQIA+ startups, thereby enriching our entrepreneurial ecosystem in its entirety.”

AI is certainly an area that could use more enrichment. Thione predicts there will be a slew of companies using AI to tackle energy consumption and distribution of power resources. “There are going to be a number of AI companies positioning themselves as ‘Hey, we are an AI company, but we’re actually using it here to solve a problem that’s going to have a positive impact on the climate.’”

TechCrunch+ recently caught up with Thione to talk about his journey in venture, the future of AI, artificial intelligence, and how Gaingels has dealt with the investor pullback.

(This interview has been edited for length and clarity.)

TechCrunch: Regarding AI, there are two schools of thought: That it will save us all or destroy us. Where do you stand and why?

Lorenzo Thione: It’s incredibly unlikely that AI will destroy humankind. I’m much more in the optimistic camp. Technology certainly has exacerbated some issues when it comes to certain types of inequalities within modern societies, but it has largely made things better. People aren’t wrong to point out the ways things can go wrong with AI.

Clearly understanding the capability of AI systems as we build them will make those issues less relevant. We are going to see problems emerge, not so much even by “AI going rogue,” but rather just the side-effects of the positive uses of AI. Those are going to have some externalities that are not positive. There will be misuse as we see AI systems go out and open source growing as a movement.

It’s inevitable that people are going to have different intentions, and some of them are going to be bad actors trying to use AI for nefarious impact. There are all of these problems, like misinformation, deep fakes, the ability to create hundreds of millions of bots, creating a distributed denial of services attack that is harder to detect, copyright, plagiarism and lack of attributions. All of these are problems that we’re going to need to contend with as a society.

AI itself is going to be the weapon to use against all this. I’m very interested in investing in companies that are effectively using AI to build adversarial systems that detect and combat the sort of misuse or side-effects of the proliferation of negative externalities of AI. Anything from deep-fake detection and bot detection to being able to create sourcing and attribution for copyrighted material. It’s a bit of an arms race, but I believe we will continue to find ways to use these technologies to make things better, not create doomsday for humanity.