Gen AI in 2025: Predictions from Alteryx Data Lead

 Alteryx's top data expert, Alan Jacobson, looks at how Gen AI will be used, to see where it's headed for businesses in 2025.

Gen AI has quickly become a big deal, changing how businesses work and make choices.

In the last two years, the base models have improved a lot, doing things we thought were impossible before.

Looking at 2025, we’re moving away from just making new models to figuring out how to use them in real life.

Businesses want to use Gen AI, so it’s important to understand the new ideas, problems, and best ways to use it. Gen AI in 2025: Predictions from Alteryx Data Lead

To see what's coming, we talked to Alan Jacobson about the future of Gen AI.

Transformative developments in Gen AI

One of the biggest changes in 2025 will be how companies use Gen AI to be more accurate and efficient.

“In 2023 and 2024, the basic Gen AI models got way better,” says Alan, “but in 2025, we'll focus on how people use these models in different industries.”

This includes making it easier for users to interact with the models and making the model's answers more accurate.

Alan also talked about LLMs (large language models) as "agents," where many LLMs work together to get the best results.

He said, “Imagine one LLM creating code, another checking its safety, another improving its speed. This will be more common as costs go down and performance gets better.”

This could change automation in areas like coding and system improvement, helping businesses improve faster and more precisely.

As companies try these new ideas, the focus will be on finding important uses for Gen AI.

Emerging use cases

Gen AI will be used more, especially in areas where LLMs do well.

Alan said the most helpful uses won't be specific to industries but will match the general abilities of LLMs.

“Summarizing large amounts of data quickly is a big use,” he says. “This includes creating insights for businesses, travel, or even politics.”

This will allow companies to get useful information from lots of data, changing how they make decisions. For example, customer service might use Gen AI to look at feedback in real time, while risk teams could sum up financial trends.

But Alan warns that Gen AI isn't a solution for everything. “Success comes from understanding what these tools are good at and not good at," he says, urging businesses to start with simple, safe uses.

Scaling Gen AI challenges

Growing the use of Gen AI comes with issues, especially around how it's managed and how people feel about it.

“It’s important for companies to ask the right questions during set up. How do we learn? How can we get experience with certain uses?” he explains.

This slow approach helps teams feel more confident and understand the technology, making it easier to accept.

Alan also talked about the problem of making people scared of AI. “Gen AI is just math,” he says. “We need to teach people how math works, what problems it can solve, and how to keep the company safe.”

By focusing on education and trust, companies can avoid rules that stop new ideas.

Looking ahead, Alan suggests IT and data teams focus on technical knowledge and making small improvements often.

“The success of GenAI isn’t about the model a company chooses, but how they use it," he says. From using LLM agents to keeping data anonymous and watching how models perform, the focus should be on how they do it.

Parting thoughts

As Gen AI gets better, 2025 will be a very important year. Since the focus is now on how to use it, businesses must learn and encourage new ideas.

By dealing with problems carefully and using Gen AI in valuable ways, companies can get the most out of it.

With good planning and education, businesses can not only handle Gen AI's difficulties but use its power to grow and succeed.