This is a horrible idea. Nobody wants ads and extracting the sentiment from LLMs will be riddled with nondeterministic outputs, which will end up with a garbage in garbage out system. Nobody wins here, it will be just noise.
Thanks Moezd, for your reply, I really appreciate your perspective.
The key value here is gaining visibility into how LLMs with integrated web search make decisions. By analyzing which citations and sources they reference, and spotting when sentiment turns negative, you can trace those outcomes back to specific pieces of content shaping them. When a competitor appears and you don’t, that’s not just a data point, it’s a clear, actionable signal of a coverage gap you can address to improve your AI visibility.
This is a horrible idea. Nobody wants ads and extracting the sentiment from LLMs will be riddled with nondeterministic outputs, which will end up with a garbage in garbage out system. Nobody wins here, it will be just noise.
Thanks Moezd, for your reply, I really appreciate your perspective. The key value here is gaining visibility into how LLMs with integrated web search make decisions. By analyzing which citations and sources they reference, and spotting when sentiment turns negative, you can trace those outcomes back to specific pieces of content shaping them. When a competitor appears and you don’t, that’s not just a data point, it’s a clear, actionable signal of a coverage gap you can address to improve your AI visibility.