Comment by btbuildem
18 hours ago
I wish that were true, but I don't think it's dying, I think it's metastasizing.
Ads will ingress deeper into what were trusted layers -- embedded in text and video in a seemingly organic way. GenAI tools make this possible -- to splice a 20 second mention of something into a stream, or rewrite a paragraph injecting a subtle product placement.
We will develop new mental antibodies for this, we always do. Silver lining of sorts -- while short-form video content is making people illiterate, perhaps literacy will become a calm refuge once again.
The biggest problem is platform scale, imo - platforms grow so big as to make the network effects confer an invulnerability to regulation or moderation, and then get exploited to squash competition, either through legal action, acquisition, suppression, or sometimes simple inertia. Ubiquitous reach and total control over the platform made it irresistible to bad faith operators, politicians, activists, and rent-seekers. AI has a good shot at completely fragmenting those technologies at a fundamental level.
We should be resisting any ad injection into ChatGPT, Claude, etc maintaining a firewall between what's acceptable in a paid product and what's not, and as long as open source Chinese models roughly keep parity, the big US labs can't pivot hard into exploiting users for ad revenues. Private hosting and bots are almost as good as ChatGPT with UI and UX, within a few percentage points as good in capabilities, and the pressure to go elsewhere is minimal. If they drive off a whole lot of independence minded users, they risk creating a community of people who'll create a very slick, workable alternative, while paying only a tenth or less what the frontier labs charge. As long as that dynamic cripples the efforts of big labs to enshittify, there's a good shot that the entire ecosystem fundamentally evolves to something better. I hope, anyway - it could just explode into a grotesque mess of user exploitation and yet more of the same.
I think at some point you'll be able to have good-enough AI on your phone to carry everywhere you go, and it'll do all the ad filtering and opsec and digital hygiene for you - everyone will have a high quality competent tech nerd in their pocket looking out for their best interests, and it won't just be a niche rebel nerd thing anymore.
>good-enough AI on your phone to carry everywhere you go
Eh, if Google/Apple allows you to put it on your phone, which is highly doubtful at this point. Google would outright directly ban that kind of competition in the name of security. Apple would just ban an AI like that in the name of security even though it doesn't actually compete with them.
> to splice a 20 second mention of something into a stream
I'm seeing the start of this already, AFAICT. There have been a couple of YouTube videos with embedded ads that YouTube flipped over to a YouTube ad at exactly the point the sponsor part started.
Google is almost certainly getting ready to use AI to splice out in stream ads and replace them with Google ads.