Moats - stronger/weaker

Now that everyone can vibe code (or "agentic code," if you want to be posh) a company of their choice....

Stronger or at least similar as pre-AI

• Network effects: Where users beget users, or where supply <> demand reinforce each other (e.g., social networks).
• Data: You can copy the app, algo, UI, marketing, landing page...but if the historical data or a real-time stream of data is what drives customer value, then there is not much you can do about it (e.g., Uber, stock exchanges, Visa/Mastercard).
• Regulatory: God-given license to operate (e.g., an arms manufacturer or a tobacco factory).
• Economies of scale: Continue to be relevant, especially for markets where a player wins by being the lowest-cost provider of a good (e.g., oil and gas, and many manufacturing and industrial companies).
• Cornered resource: Similar to pre-AI (e.g., you own the only highway in and out of a region).
• Brand: Was always meh, and don't see it changing much in AI

Weaker:

• Switching costs: A lot of the migration headaches of moving OUT of a piece of software have significantly reduced. Bear in mind that if there are OTHER moats around that piece of software then those will still continue to be valuable.
• UI/workflow patterns and process power: You have been trained to use equipment/software X for years, and so using a competitor is harder. Well, f**k that, because you will be operating less of it directly in the coming years (AI agents will).