I love the concept of “frontier firms” that are pushing the edge with AI. They are undoubtedly getting ahead in their respective industries.

In 2026, frontier firms are going to keep things interesting. A lot of these firms will start to hit a ceiling with what they can do with their current approach, like API wrappers and fine tuning of models.

They will overcome this ceiling by focusing more on foundational models. Models that are built-for-purpose, from the ground up, for a particular industry or even a specific company. These are models that will take them to a completely new level in their adoption of AI and in their industry.

To be sure, a lot of companies out there currently claim to have a foundational model. But don’t be fooled - they don’t. The rise of foundational models will really hit a stride in 2026.

It would be very hard for you to go from being a student or beginner data scientist and into developing foundational models. That’s a long road that requires some skill development.

But if you’re already a professional data scientist, looking for the next meta-trend, you could start preparing to be an expert in foundational models. There will be increased demand this year for this role, I guarantee it.

Your data science book reading list on Amazon

When you’re ready, here’s two ways I can help you:

- Join other data pros getting exclusive tutorials and weekly deep dives: https://clients.datasciencewithdennis.com/upgrade

- 1-on-1 Strategy: Book a 60-minute session to map out your data career or AI project: https://calendly.com/dswithdennis/career-consultation-with-dennis

What’s The Latest?

What I’m Loving Right Now

World’s First Safe AI-Native Browser

AI should work for you, not the other way around. Norton Neo is the world's first safe AI-native browser with context-aware AI, built-in privacy, and configurable memory. Zero-prompt productivity that actually works.

Keep Reading