Vertical Agents

It's not ideal for small to medium size stage businesses to have 20 different AI verticals: one for marketing, another for research, communications etc. It starts to become expensive quickly.

If you're focusing on a vertical, you must understand that you're competing with other verticals, not in your industry but also in other industries. The decision for a business to spend money on your vertical then becomes a matter of what important problem they are facing at the moment. And anyone who's started a business knows that changes often. For the pre-AI saas era, that was ok, but now we have models that can generally solve most business cases, even if vague, the solution is still better than paying $200 for each of 5 different AI saas verticals.

A common argument is to focus on an area that the models don't perform well on. This may be a good strategy for today's exponential buildout phase. But as we advance further up the scale, the models perform better on areas they did not previously perform on. This is the bitter lesson for verticals.