> The results you see in OpenAI searches are not paid placements, but organic results. “They are not ads,” says Fry. “They are not sponsored.”
We all know that won't be the case forever. But I'm also pessimistically interested in the modern SEO tactics of optimizing your product to appear in model outputs.
AndrewKemendo 2 hours ago [-]
It’s the same approach as always: Capture the reference data as close to the source as possible
So any standard around RAG will be a place to attach the adserver
Any UI will catch the semantics in the metadata analysis in order to parallel construct in some sense the affinity graphs
So there’s always leaky personal targeting data - with enough users you don’t even need to provide much data to fall into a targeted group
nicce 1 hours ago [-]
Meta and Google are one of the most valuable companies while people say that they are immune to ads and cannot be influenced.
kylehotchkiss 4 hours ago [-]
I imagine this more as you pay ChatGPT to feature your product or more likely, to silently add your product to somebodies cart
nitwit005 4 hours ago [-]
I hadn't considered SEO-like tactics. I do expect some gen-AI companies to let people pay to insert ads or marketing, because it can be as simple as tweaking a prompt. Have your art generator insert Pepsi Co products whenever beverage terms are used.
sethhochberg 4 hours ago [-]
I can certainly see the potential: both for this to be really useful at the start and progressively move toward the useless as someone figures out they can get a big bonus if they sell sponsorship within it.
This last weekend I was using ChatGPT to help me select complementary neutral and trim colors for a room where I’d already picked out the accent wall color and other decor, because I got tired of wading through SEO-spam interior design blogs. I gave it quite a bit of detail about my general style, lighting conditions, even things like the various woods used in my flooring and furniture so it could be conscious of complementary undertones.
It would’ve been convenient if once I had the list from the conversation, I could’ve asked it to just order the sample paint chips from Benjamin Moore and have them sent to me. And by that point it knew quite a bit about the size and features of the room, maybe it could have added some requisite supplies to my Home Depot cart for pickup…
The real value is a tool that can act on all the conversational context in a relatively vendor agnostic way. Take my entire plan and use various agents to make it happen.
xdfgh1112 5 hours ago [-]
Google is corrupted with over half the search results being related to buying something. I search for regular words and it's interpreted as a brand or product.
If that happens to ChatGPT, it's useless to me.
I suspect the free version will tilt more in this direction until it becomes worse than useless. The current free offering is honestly too good to be true at present.
beAbU 4 hours ago [-]
Ironically, these days I only switch from ddg to google when I want to buy something.
mk89 6 hours ago [-]
It took them less than I thought.
We are gonna have fun with "can you suggest me a product to do XYZ"?
At least until now you could run a search to compare, etc. ...Imagine how biased the results can be when you ask "show the pro/con of each".
Less and less trustworthy.
bluedevilzn 5 hours ago [-]
It begins...
Imnimo 5 hours ago [-]
>The results you see in OpenAI searches are not paid placements, but organic results. “They are not ads,” says Fry. “They are not sponsored.”
What does ChatGPT use to do web searches? Is it Bing with sponsored results removed?
isubkhankulov 2 hours ago [-]
Yes that seems to be the case. perplexity uses Google results.
caseyy 4 hours ago [-]
> The results you see in OpenAI searches are not paid placements, but organic results.
Aaaaaaand it’s gone. Please step aside, this chat is for people who will actually buy the sponsored products. Next please!
4 hours ago [-]
SirMaster 5 hours ago [-]
Wake me up when ChatGPT can buy me a new nvidia 50 series GPU at MSRP.
We all know that won't be the case forever. But I'm also pessimistically interested in the modern SEO tactics of optimizing your product to appear in model outputs.
So any standard around RAG will be a place to attach the adserver
Any UI will catch the semantics in the metadata analysis in order to parallel construct in some sense the affinity graphs
So there’s always leaky personal targeting data - with enough users you don’t even need to provide much data to fall into a targeted group
This last weekend I was using ChatGPT to help me select complementary neutral and trim colors for a room where I’d already picked out the accent wall color and other decor, because I got tired of wading through SEO-spam interior design blogs. I gave it quite a bit of detail about my general style, lighting conditions, even things like the various woods used in my flooring and furniture so it could be conscious of complementary undertones.
It would’ve been convenient if once I had the list from the conversation, I could’ve asked it to just order the sample paint chips from Benjamin Moore and have them sent to me. And by that point it knew quite a bit about the size and features of the room, maybe it could have added some requisite supplies to my Home Depot cart for pickup…
The real value is a tool that can act on all the conversational context in a relatively vendor agnostic way. Take my entire plan and use various agents to make it happen.
If that happens to ChatGPT, it's useless to me.
I suspect the free version will tilt more in this direction until it becomes worse than useless. The current free offering is honestly too good to be true at present.
We are gonna have fun with "can you suggest me a product to do XYZ"?
At least until now you could run a search to compare, etc. ...Imagine how biased the results can be when you ask "show the pro/con of each".
Less and less trustworthy.
What does ChatGPT use to do web searches? Is it Bing with sponsored results removed?
Aaaaaaand it’s gone. Please step aside, this chat is for people who will actually buy the sponsored products. Next please!