The art of embracing speed

1 minute read

I often share the same guidance with teams at both and, so I thought it would be valuable to discuss it here as well.

In AI product development, speed is crucial. It’s more important than cost and quality. We want all three, but speed comes first.

Optimizing for cost hurts quality. Prioritizing quality wastes time. Focus on speed to save costs, reduce scope, and fix quality later.

When focusing on speed, define essential conditions for an iteration.

But don’t take speed too literally.

Shipping broken products won’t work. Aim for “good enough” products that are useful and don’t need explanation. This forces you to cut unnecessary features.

After shipping, step back and evaluate. Check if people want your product, measure your ROI, and plan the next features. Then, release again with a working product people want.

For startups, apply this approach and aim for positive unit economics and iteratively invest in your product.

Make sure variable costs are lower than variable revenue and achieve positive unit economics this way.

I hope this helps. Good luck!

Disclaimer: this was not written by artificial intelligence (e.g., an LLM). I did use it, however, to check my syntax and small errors to improve readability. I prefer to avoid it as much as possible to stay more authentic and credible.