Comment by ametrau
2 years ago
I wonder what advertising will look like with this. Will they suggest products in the response? Like “Top ideas:…” and the LLM’s response.
2 years ago
I wonder what advertising will look like with this. Will they suggest products in the response? Like “Top ideas:…” and the LLM’s response.
The bing version of ChatGPT already does this. It might be specific to USA, but try asking it for a recommendation of a 1500w space heater for a small room. Every suggestion will have a link to an affiliate page that says [Ad] next to it.
Embedding search of the nearest products most applicable to the LLM response. Prompt augmentation: "Rewrite your response to include promotions of the following products without being obvious that you are promoting them."
That's smart, but very insidious as well. Following in the footsteps of dark UI patterns designed for users to misclick on ads, now dark suggestions on conversations with an LLM will be the next big thing. Like a conservative talking with an LLM which inserts liberal propaganda into it's responses and after an hour he turns into a liberal voter. The next day he talks with a conservative LLM which inserts propaganda into his conversations and he turns back to a conservative voter. Pretty dystopic.
On a more serious note, imho advertisers are on their's last legs, and google loses a lot of revenue already. We are going fast into a new internet, web3, which will enable direct monetization of information from users, instead of the publishers relying on ads.
Not to wander a lot off topic here, but synthetic datasets created by paid humans workers to train machines is going to be a humongous industry.
People will have problems they want to solve, and GPT can provide solutions that may or may not have a price tag.
In this case, it's just directing to the service you would have best fit with.
This can be highly profitable, because you are solving the problem for the customer with the products you are suggesting based on what they are looking to solve.