Fyne.io in Go also doesn't want to support a browser, but that is a great way to expand mindshare (similar to implementing a solid data grid in a UI toolkit).
Why can't anyone just run it through one of the "make this not appear so clearly LLM-generated" obfuscation utilities open sourced on GitHub, or something similar just with prompts.
The level of phoning in the complete lack of any work beyond the absolute bare minimum is the most frustrating aspect to me.
I would say, because not everyone cares as deeply as you do.
If it's readable and aesthetic, it is perfectly serviceable for the purpose of reading a post.
Consider that you probably didn't have such an intense response to WIX templates or substack articles, so most of what's going on is actually your feelings about AI taking primacy.
> Consider that you probably didn't have such an intense response to WIX templates or substack articles
Apologism at best. People have been bemoaning this issue for a while now..., pre-AI even.
First time I'm seeing this defended from an "oh you're just anti AI" perspective (which, ironically, is exactly the sort of bare minimum a couple comments over was bemoaning)
If you create content that people like, with a distinct style, one can direct their LLM to it and copy the style. The more content you have out there the better the copy will be. I don’t think it’s really a differentiation you can rely upon
The world you are describing assumes nobody will produce anything anymore because of theft. I think there will be a reset and hopefully content producers like Mr Beast will lose out. But, I do not think people, similar to the original participants on the Internet, will refuse to make anything. I expect a return to Live journal similar content where we get actual opinions even if we sadly lose out on long tail content/production because people do not want to work hard for free. The return of the hobby style Internet isn't necessarily bad.
> The world you are describing assumes nobody will produce anything anymore because of theft.
I don’t believe that at all, and it’s not implied in my comments. What I’m saying is that the value isn’t in the style, that part can also be replicated. It’s not in the content, that can be replicated. It’s the same for >90% of startup ideas people come up with when discovering agent coding. You were able to create a cool SaaS project in a weekend? Congrats. Any other dev can point their LLM to your website/app and make a clone in even less time. The safe ones have a high barrier of entry, such as hardware, or similar, but the vast majority is trivially cloneable.
I don’t know what that will result in. I assume we will be chasing more human experiences, but I have no idea.
In the spirit of not speedrunning towards the dead internet theory, then what we need is a slop-detector that detects things that have been run through a script to "make this obviously not look like slop", in a recursive looping battle royale until the heat death of the universe.
Cloudflare failed pre-AI with Cloudbleed, which could have been alright except for how they initially handled it. In addition, a few checks & balances have retired or stepped back so there's nothing stopping AI amplification of that company culture.
They may succeed or not, but right now it's hard to stomach as a former fan. I see where a blog post like this could be coming from.
loaned from German, where it's originally a way to say buttocks, literally "sitting flesh". If you have more Sitzfleisch you can sit for longer. Both in the literal sense (a bigger butt makes sitting more comfortable) and in the figurative sense (having the mental ability to sit for longer, get more desk work done)
I think there is an IT bit of humor from about 1 decade+ back where the people who's proposals won out in meeting were the ones that could keep from needing to go to the bathroom longer.
Doesn't this just mean criminal suspects will be forced to unlock their devices using biometrics (in the US)? That should buy law enforcement some more time until biometrics go out of favor.
For the HN 14 day record since I didn't see this yet: in genuinely critical situations, if the threat model includes a device/service being used against one's self, "the only winning move is not to play" (stop the use of the device/service!)
In the same manner, if the critical concern is a device/service being used against one's self remotely, stop the use of the device/service if it is capable of remote access or data collection for later retrieval.
The ultimate baseline reality is that deciding to care whether or not remote access/data retrieval is using a method supported by the device manufacturer/service provider and/or whether doing so is allowed by the current legal system is too late.
Premptive HN disclaimers: I acknowledge the growing unavoidability of devices and services supporting remote access and/or data collection as I type this here on my phone. Remote access and data collection are very convenient and somehow still expected by device owners and service users to be under their control alone. It can be difficult to determine if a device or service supports remote access/data collection. There are often immediate consequences and eventual legal consequences for stopping the use of devices or services that belong to others. Working to change the legal system to slow/reduce use of devices and services against one's self and encouraging others to do the same is admirable.
reply