Some people write culture pages, some philosophy. Others work on humble about pages, and many keep it for the careers page. We think it’s more useful to capture the key aspects of how we think.
AI Native
We’re living through a fundamental shift in how work gets done. The confluence scaled hardware, AI models, and growth of workloads such as knowledge work and research is fertile ground for innovation across all industries and design domains.
Pomona is an AI-native company – which means we seek to offload as much as we can to AI systems and have built our team, processes, and tools around these capabilities. GPTs aren’t cool new toys to enhance our workflows, they are foundational to our organizational design.
Intelligence by design
Most software simply captures rulesets and data transforms and offloads the work of inference to human operators. This model is dead. Software can now do much of this inference at scale, so all solutions must be designed with this in mind.
Inherent chaos
The world is complex, but most software pretends this isn’t the case. We simplify until it can fit into a web request or database table, but now we can chomp right through a lot more complexity. What happens when we broaden our inputs and outputs?
Maximize potential
The beauty of more intelligent solutions is that we can both scale up our dreams and work on the highest-impact problems. By offloading toil and increasing our throughput of complex workloads, we can focus on the stuff that really matters.
Non-determinism
Our world is continuous not discrete, but we’ve only ever worked on discrete systems at scale. We’ve encoded knowledge into static software, but now we have the opportunity to get comfortable with more complexity and multiple right answers. Our software can now give us finger in the air estimates and confidence intervals for questions we pose.
Definitive business impact
We’ve built all modern AI tooling as goal-directed agents, which means they’re extremely well-suited to delivering business impact. Directed correctly we believe baking intelligence into our workflows and solutions will power the next economic boom (both directly, and indirectly through freeing people up to work on non-AI-powered problem domains.)
Human delight
All of this is pointless unless we’re actually improving human experience. Profits are lovely, but our collective moment-to-moment experience is more important. We should build systems that enable maximum human delight.