OPTO Sessions: How AI Transformed Business in 2025

2025 was the year that artificial intelligence (AI) consolidated its hold on global business. 

Throughout the year, OPTO Sessions spoke to leaders in sectors as diverse as lending, customer service and cloud storage. Here’s what they had to say about how AI is impacting their businesses and reshaping global markets. 

Viable Use Cases

A completely autonomous customer service agent. A 24/7 mortgage advisor. A smart voice assistant built into your car’s dashboard. Fast, AI-approved loans. 

These are just some of the revolutionary AI use cases our interviewees talked about over the course of 2025. 

What are the concrete benefits consumers and businesses can expect from AI? 

While much of the talk around AI has centered on efficiency gains for companies, advancements in agentic AI can represent real wins for customers, as Leah Price, head of the Tinman AI platform at Better Home & Finance [BETR], explains. “If I’m a borrower, if I work nine to five, I can’t use a user interface… soon you’ll be able to speak to one of our avatars in the middle of the night. You’re nervous about your mortgage application. You want to have a FaceTime call with an expert mortgage advisor. You can do that.”

These benefits also extend to client businesses. Des Traynor, Co-Founder and Chief Strategy Officer at Intercom, envisions Intercom’s Fin AI becoming an agent capable of learning and adapting to internal changes within a company. “Our goal with Fin is that the human support team should be answering things for the first time and the last time.”

For Paul Gu, Co-Founder at AI lender Upstart [UPST], the win is in the time saved. “Why do [people] care about credit? It’s because one day they hope to be a homeowner. They hope to be able to buy a house when they need it. And probably they hope to be able to get the loan quick enough that they can actually close on the home that they’re trying to buy instead of running out [of time] or someone else grabbing the house.”

Add in voice as an interface for AI and the use cases are limitless, SoundHound’s [SOUN] Chief Financial Officer Nitesh Sharan attests. “Our vision is to voice-enable the world with conversational intelligence. So that doesn’t mean it’s unique to any one industry … we’re trying to enable transactions, commerce and discovery through all the ways consumers want to discover.” 

Paying the Bills

While the potential uses for AI can be easy enough to determine, some skeptics may ask where the money is going to be made, long term. 

Nowadays, chipmakers like Nvidia [NVDA] and applications like OpenAI get most of the press; however, Box’s [BOX] CEO Aaron Levie argues that the value is starting to trickle down to a wider range of companies using AI. “At the application layer, we’re finally at the point where enterprises can utilize AI across their business processes. I think you’re going to see that value flow through at the software layer.”

With many onlookers worried about a potential AI bubble, it can be helpful to consider how continued advances will increase the value AI brings to businesses. 

In terms of customer service, the focus is simple: “proactive support”, Intercom’s Traynor says. “How we can get ahead of problems for our users versus how we can detect a struggling user, somebody who’s misconfigured, somebody who’s about to have a problem. Every time you’re releasing a product and you have a red blinking light or you pop an error message, you should really assume that is the beginning of a support conversation.”

Upstart’s Gu is optimistic about his company’s role in applying AI to consumer credit. “In five years’ time, I think we’ll have successfully brought AI lending to the parts of the consumer credit market that actually are big and matter to people,” he says, namely: auto, home and revolving credit. 

“When that happens, I think there’s going to be a seismic move in how people perceive the relevance of AI and lending today.”

And that could mean more equitable credit access for more people. “The end result of that is you’re actually going to asymptotically approach the real levels of access to credit that should exist in this country. Something like 80% of Americans should have access to bank quality credit. They should have it on demand, with virtually no work required of them. And the other 20% should have a clear and defined path of how to get from where they are to where they want to be.”

Box’s Levie emphasizes that behavioral change can be an even greater challenge than technological progress. “It’s going to be a multi-year journey. The change management in a large enterprise is non-trivial, getting people to use a new set of tools in a new behavior.”

New behaviors, and not just new tools, he says, are part of what is making the AI revolution so revolutionary. “For a lot of the past 10 to 20 years of software, we didn’t really have behavior change. We just had tool change,” he explains. With AI, “what we have here is behavior change… the way I actually interact with software is totally different. I’m no longer doing all of the work myself within software. I’m actually interacting with AI labor on the other end.”

Sorting Through the Slop

It is telling that the Merriam-Webster Dictionary chose “AI slop” as its word of the year. While this cutting-edge technology really hit its stride in 2025, the hype surrounding it — and the billions of dollars being invested in AI-focused computing and infrastructure support — have made it hard to tell what’s useful and what’s slop.

Leah Price’s background in the public sector has given her experience determining the real value of an AI solution. “Basically any company can go out there and make claims, oh, we have all this AI, it’s really driving our efficiency and it’s really cool and show off some chat bot that they’ve slapped on top. And there they are, they’re an AI company. They should trade at a higher multiple and they can get a lot of attention.”

So how should investors figure out where the real AI innovators are? 

“You need to do the homework yourself.” 

While it can be easy to be impressed by the numbers, “do not only listen to an earnings call,” Price insists. Instead, try out the AI solutions a company touts to see if they really work. “Push our AI hard. Ask her hard questions,” she recommends for Better’s Betsy AI chatbot. “Then I want you to go to some other company websites … Do a side-by-side comparison. See how far you can push it yourself.”

In the end, more discerning consumers will necessitate better-performing AI solutions, Price explains. 

“If I can get my competitors out there feeling like it’s a little bit uncomfortable, that’s when I know I’ve done my job,” she says.

Intercom’s Traynor agrees. “AI is rife with what I call shelf-wear,” he explains. “It was bought, it’s sitting on a shelf, no one’s touched it for a year.” As an investor himself, he has outlined four criteria to determine if a so-called AI-first firm is the real deal. 

First, “it’s really important the product is used by customers.”

Second, it should solve a real business problem, rather than being what he calls a “Cool Tool”. 

Third: “There needs to be deeply differentiated AI… It can’t be something that I could build in a weekend by just handing the whole problem over to OpenAI or Anthropic.”

Lastly, and most importantly, “there just has to be some evidence that the margin can turn positive… It’s really important that you’re not spending a phenomenal amount of tokens to do something that is not of proportionate value at all.”

In the end, it all comes down to good business. “If you’re burning the entire haystack to find the single needle and people don’t even really care that much about the needle, that’s not an efficient business.”