Product teams that have an easier time integrating AI into their products have the mindset: "No one cares if we use an LLM or not - what is the experience we are aiming for"

With these labs all outdoing each other it seems there is very little to gain from simply training large language models.

Now, shipping better software and applying these language models that solves a problem 10x better than rule-based solutions?

That will be where the money lies.

There is nothing quite like the adrenalin rush of deploying database changes in to production 😅

#software #engineering

As a software engineer working on SaaS products and a researcher, I believe that learning about new technology is the best way to deal with these types of waves.

Staying curious, not only on an individual level, but as a team is a huge advantage.

The big take away from all of this DeepSeek news is this: Language models are still incredibly new and putting all of your eggs in one basket could cost you valuable developer resources.

Stay flexible and build a robust team that can execute on changes quickly.

As we are heading into the final month of 2024 I felt incredibly grateful to be working on delivering value through the use of technology.

Designing for AI for me always meant to have one foot in the future and one in the present.

Can't wait to see what 2025 will bring ☺️