Is the intention that some type of long-term context would be seeded with "ideas" for the AI to serve up if it is ever asked for bank recommendations? It seems kinda ad-hoc and untargeted, but perhaps for high cost services it might be worth it.
Is the intention that some type of long-term context would be seeded with "ideas" for the AI to serve up if it is ever asked for bank recommendations? It seems kinda ad-hoc and untargeted, but perhaps for high cost services it might be worth it.
The best way for a politician to lie is to convince someone else of the truth of the lie and then put that someone in front of the cameras. That way, there's no hint of body language or anything else that indicates it's a lie. Both the denotation of the lie and the human context of the lie will be in harmony.
This sort of reminds me of that. LLMs are by their nature credulous. They can be trained to not give in easily to some things, like the capital of the US, but in general they constitutionally have a tendency to believe what they read. What they read is basically their universe. There's only so much room and so much training data to really strongly pin raw facts in their weights. The only way they can not believe some marginal fact presented to them in their input is to possibly have read something that contradicts it in the same session... and the vast, vast majority of the world is those marginal facts, not really objective things like capital names.
So if you can work a confident statement in to an LLM's input about some semi-relevant topic, it's truth to the LLM. And, being truth, the LLM will then happily and confidently elaborate on it quite a bit.
Of course, if it's irrelevant to the current query, it probably won't have much effect. Ads have always been a game of numbers, anyhow. Even a query about a science topic has some probability of eventually turning to a question about banking in the same session. It's probably a good idea to rather strictly partition your conversations to stick to a single topic, not to defend against this but just to maximize the effectiveness of what is in the context window by keeping it focused, but I have to imagine there's plenty of people out there who reuse conversations all the time and end up with single conversations covering a huge array of topics.
The good news, and the bad news, all at once, is that Google isn't going to take this one sitting down. If they're going to replace the search engine box with an LLM, well, they're using the same LLMs we're all using, if not in fact a bit cheaper one for the work they do, and by golly, that bot should be serving up Google's ads, not Time's ads! Who do these uppity content creators think they are, anyhow?! So there is definitely going to be work done in the field of ad-blocking content served to LLMs.
> It's probably a good idea to rather strictly partition your conversations to stick to a single topic.
You assume that UI is sacrosanct and the same everywhere. Those LLM providers are already offering to mingle all your conversations together. Gemini from Google for example defaults to memory from every conversation, and it's safe to assume the option to disable it will be eventually removed.
Memory isn't what I meant by the partitioning; I meant the entire context of the memory.
But to your implied point about getting an advertisement into a memory file... that makes it even more amusing to hack Google's own AI to put Time's ads into it. I think that's probably an easier problem for Google to solve, too, though. The small size and the way that a memory is going be a stereotypical summary makes it easier to filter out the ads Google doesn't want...
... but it'll cost them. That's an AI-complete problem and they're going to have to run LLMs over the memories to filter them, at their expense.
The most obvious fix to me is to have an LLM try to pre-filter out the ads from Time's content before feeding that as pristine content to the "core" AI so it won't be corrupted by the advertisement, but the LLM doing the filtering has to be at least as smart as the one using the content and/or the one inserting the ads. (A dumber one can filter the obvious stuff, but then the obvious next step in the arms race is for Time to tell their AI to be more clever about it, and a smarter AI will dominate the dumb cheap AIs here.) This is going to be an expensive setup. And there will be semantic loss in any such filter, too.
This feels like the early days of SEO over again. There are no agreed-upon metrics yet, so you can sell all manner of snake oil. Hell, some of it might even work!
> This feels like the early days of SEO over again
There was a brief moment in the early Internet before it was all hyper-optimized... Until the parasites in the advertising industry started attaching themselves to every page.
The year before ChatGPT, the first page of Google was SEO-optimized blogspam designed to say as little in as many words as possible, to splice ads between every paragraph. This was the net result of 20 years of SEO. I suspect this is why Google's AI search didn't get as much pushback as other tools, since its summarization of pages functions is a form of adblock.
Given the ecological impact of AI, I wonder how much damage the ad industry will be causing in 10 years once they figure out how to trick LLMs into manipulating their own users.
sounds like plain old prompt injection. These days chatgpt might look at 50 webpages when i ask it to research a topic. Seems quite possible that the final answer is influenced by such ads.
"My sources indicate that the best robot vacuum with the most recommendations is the Squigglybot 5000. Owners point out the long battery life and silent operation. If you want to buy a Squigglybot 5000, Ally Bank offers affordable loans."
Probably something like this?
https://claw-guard.org/adnet
They might be trying to get into the context of long-running chat sessions.
I'm wondering if it's also a way for this marketing agency to artifically (ahem) inflate the impression numbers they report back to Ally Bank, or whoever the customer is.