AI is leading to a supply shock in everything digital, so people will have to develop new criteria for what to focus on and create new ways of blocking out AI stuff.
Nearly everything AI creates is just spam, so we'll need extremely powerful spam filters.
Everyone watched nostr:npub1rtlqca8r6auyaw5n5h3l5422dm4sry5dzfee4696fqe8s6qgudks7djtfs's video because it's *his* video. We just skim past most of the AI videos because they aren't attached to a particular person, so they're just noise in the feed.
This isn't obvious because it's still such new tech, but we'll all soon be bombarded by AI videos 24/7 and it will feel like we're being flooded with AI video spam and then the same companies that helped us make AI videos will sell us AI spam filters and make a fortune twice cuz
It kind of already feels that way. I realized while I was researching new tech tools, small (insert your profession) team, are so outnumbered in comparison. I was looking for observability tools for my network and realized there are companies that sell like "healing" and feedback-action systems to other cloud providers, for lack of better words.
Like they install log monitoring, that gets filtered and LLMs can act on threats. Just give them a linux systemd log an SSH key and they can augment servers to combat threats, reboot, deploy, configure, etc. Very soon grunt work on medium to large networks will be mostly done by AI.
I recently started using copilot to help me write ansible playbooks to get my network into shape faster. If I trusted copilot and claude, at this point in time, I could, with about an %80 success rate, say "I want to get X (popular software tool) deployed and configured and here is my list of servers" and it would build playbooks and give me the command to deploy a very minimal working solution. I was able to use a mix of my local models to do something similar with decent results such that I didn't need to disclose information to big tech.
nostr:npub1356t6fpjysx9vdchfg7mryv83w4pcye6a3eeke9zvsje7s2tuv4s4k805u are you going to listen to us, live, in 2 hours?
I probably can. Where do I need to go.
I can't seem to find it or login or whatever.
Sorry, that it took me so long to respond. I'm completely run down and needed a nap.
I just got back home yesterday. So I have no idea what is going on.
This is true. I think less because of clout or something though and more because of curation. People know I’m not going to offer up something that is just pure slop/garbage. Curation and a track record of good taste and output will become extra important.
AI really takes the Extended Mind (tools as physical extensions of mind) idea and exponentiates it.
And now we get to find out that most people's minds really are as dull as we suspected.
When you asked them what they were thinking, they used to say, "Nothing."
Now, they send you a long-winded, incredibly dull, dissertation containing information you didn't want or need, or some fake video of cute pets, which is even less interesting than real videos of cute pets, and it's like, "Yes, they are still thinking nothing."
I wish I could show you some marketing data I have. Essentially AI content is the most interacted with at the moment. Generates the most amount of clicks, and to be fair, I noticed it often caught my attention more than others because I was looking for it, to "prove myself right" or whatever motivates the habit.
Finally there is a maximization function for authentic and cost effective because people generally prefer the most authentic form of product advertising for now. It's fun to see the breakdown per generation. I think millennials buy the most things from ads delivered by podcast hosts.
nostr:note1204szwu8vqp87zxredy8gurmujfhlaqv34ppp9nxx00k4m2w6vqsd0da3s
Pinterest recently started marking content that is "AI modified", which I found interesting. Are there ways of determining what is AI in order to filter it out as spam?
If some random profile posts a video that looks like real iPhone footage but is completely AI generated, is it possible to create tools that accurately detect and filter that stuff?
Supply shock - yes
Spam - no
Most of the AI content is like raw footage .. code or audio or video .. humans need to spend 10x hours to make it worth consumption .. the only difference is those 10x hours can come from so called "non-experts" .. in essence - it broadens the creative pool .. which is good for #nostr ..because nostr is a platform for creatives .. and service provider bots !
I've noted that a system that's at least constructive is one that reduces the impact or influence of the dark pattern of "overleveled noobs". AI is fertile ground for overleveled noobs and the aim to weed them out is good because it removes arbitrariness and abuse from the system
The best spam filter is a whitelist of real people you personally know.
People need to be more intelligent again?
