I like this & really agree. I sometimes wonder though, if the goal is to abstract most of the friction so the user just experiences a magical outcome, at what point does the product become infra / lose the end user's touchpoint (and thus mindshare)?
I think it depends what you mean by mindshare, if by that you mean a person then yes I agree that we could lose that. But isn't what Elena is arguing for is redefining the goal as increased "agent share"?
We are slipping into a new, as yet undefined form of valuation for the use of software.
My prediction: Claude ain't gonna keep charging $20/month forever after their IPO. At some point the pricing for 'generalized AI software' and specific, narrow tools will find alignment.
Thanks for covering this important topic! I completely agree that trust is a more important barrier now when it comes to uptake. And the uncertainty aversion that comes with AI use, and not knowing the best way to use it.
I think agent usage brings a whole class of problems that PMs will need to come to terms with. Not only getting an agent to find and use a tool, but also ensuring it chooses the right tool. My guess is an agent using the wrong tool will hurt trust more as the user won't know why the output was not as expected. Curious how you keep track of that at Lovable?
The Slack scene over your agent not knowing a feature has an outside version. ChatGPT doesn't know it either, and there's no channel to yell in. It goes by what's public. So use case discovery starts before your prompt box, with whatever a buyer got told when they asked what you can do.
Strong framing, Elena. I think this lands cleanly for single-player and B2C products, where intent lives in one head. B2B is different. The “what do I want to do” varies by persona, by task, and by what the org is optimizing for this quarter, and all three keep shifting. Predefined workflows were never just a discovery workaround in that world. They are how companies get consistency across hundreds of people and make roles transferable when someone moves on. My bet is that B2B doesn’t drop workflows; the agent becomes the thing that executes and adapts them within guardrails the org defines. The discovery problem shifts from “which feature?” to “whose intent wins, and who governs it?” Curious whether you’re seeing that at Lovable’s team accounts.
You are talking about trust, trust of people and of agents in the software we market.It seems to me that the role of brand building is increasing again. What would you say?
I like this & really agree. I sometimes wonder though, if the goal is to abstract most of the friction so the user just experiences a magical outcome, at what point does the product become infra / lose the end user's touchpoint (and thus mindshare)?
I think it depends what you mean by mindshare, if by that you mean a person then yes I agree that we could lose that. But isn't what Elena is arguing for is redefining the goal as increased "agent share"?
Very quickly. The term is "commoditization".
We are slipping into a new, as yet undefined form of valuation for the use of software.
My prediction: Claude ain't gonna keep charging $20/month forever after their IPO. At some point the pricing for 'generalized AI software' and specific, narrow tools will find alignment.
Completely spot on. I just wrote about it too. https://charliegautreaux.substack.com/p/why-your-next-customer-may-never
AI is the capability - not a feature. We will talk in terms of capabilities, features are delivered dynamically and in real time.
Thanks for covering this important topic! I completely agree that trust is a more important barrier now when it comes to uptake. And the uncertainty aversion that comes with AI use, and not knowing the best way to use it.
I think agent usage brings a whole class of problems that PMs will need to come to terms with. Not only getting an agent to find and use a tool, but also ensuring it chooses the right tool. My guess is an agent using the wrong tool will hurt trust more as the user won't know why the output was not as expected. Curious how you keep track of that at Lovable?
The Slack scene over your agent not knowing a feature has an outside version. ChatGPT doesn't know it either, and there's no channel to yell in. It goes by what's public. So use case discovery starts before your prompt box, with whatever a buyer got told when they asked what you can do.
Strong framing, Elena. I think this lands cleanly for single-player and B2C products, where intent lives in one head. B2B is different. The “what do I want to do” varies by persona, by task, and by what the org is optimizing for this quarter, and all three keep shifting. Predefined workflows were never just a discovery workaround in that world. They are how companies get consistency across hundreds of people and make roles transferable when someone moves on. My bet is that B2B doesn’t drop workflows; the agent becomes the thing that executes and adapts them within guardrails the org defines. The discovery problem shifts from “which feature?” to “whose intent wins, and who governs it?” Curious whether you’re seeing that at Lovable’s team accounts.
great article
You are talking about trust, trust of people and of agents in the software we market.It seems to me that the role of brand building is increasing again. What would you say?