A software developer and Linux nerd, living in Germany. I'm usually a chill dude but my online persona doesn't always reflect my true personality. Take what I say with a grain of salt, I usually try to be nice and give good advice, though.
I'm into Free Software, selfhosting, microcontrollers and electronics, freedom, privacy and the usual stuff. And a few select other random things as well.
I think it's basically what you said. There's some organizations on Mastodon, some German newspapers and tech magazines, game developers, a few artists and some more known people from hacker spaces. We don't really have that here.
I wonder, though, are you a human or an AI agent? I don't think you replied to kata1yst. And you wrote another reply which doesn't really fit what you replied to. You might have the AI equivalent of "fat finger syndrome". But I can't give good advice unless I know what kind of entity I'm talking to...
Awesome. I'm still looking for some open-source solution which does pretty much what you do, plus a graphical avatar. I don't have an exact use-case but I imagine it'd be nice as an AI Vtuber, maybe some virtual creature (I got a few unused displays for ESP32s in my drawer) or some sort of desktop assistant / secretary.
What I currently do is use ESPHome and Home Assistant. I got a few satellites around the house and they can do speech commands like change the light, turn on the TV or set a timer or inform me about the weather conditions... Unfortunately I'm lacking a GPU in my server so all of that is a bit more dumbed down to keep the latency down. So no tool calling or LLMs or any fancy stuff. Requires me to say fixed commands, but in turn makes it incredibly fast, even on a CPU. It'll do for smarthome application, but I suppose that's a different scope from what you're programming.
Btw, just to let you know, in case your "computer" model is still to be trained... (as you write in the Readme) There's a bunch of people who make microwakeword models for ESPHome and put them on GitHub:
How much rate-limiting do you get with a VPN? Because I use SearxNG with my regular network and I tend to get a lot of my search queries fail because the search providers have CAPTCHAs after the 10th request or so for some given timespan.
Yeah, I think it's a bit sad it has to be like this... But it is what it is... I once moved and asked my neighbours if I could use their Wifi for a month, and that was basically their answer. Just that I wasn't some internet stranger, but a regular stranger living nextdoors.
By the way, I'm not sure whether your posts fit here.
I think it'd be better if you wrote one summed up blog post / study result. You don't need to keep us up to date every day with what you asked AI and what it got wrong... I mean don't me wrong, either. It's great and all how people study AI... But nothing here comes as a surprise to us. I think by now, pretty much all of humanity knows how AI makes a lot of mistakes. That's not really a groundbreaking result. Also not directly related to open-weight models or Free Software.
But could also be me and I'm not really aligned with the rest of this community, dunno... Do other people like the posts? Because I'm not sure if I should downvote or not. I'd rather talk about AI, or make my own chats. It's just, rarely do I feel the urge to work through other people's chatlogs... Unless that's followed up with 3 pages of maths and conclusions...
Also: Didn't you miss A LOT of mistakes in the chat? They go on and on discussing nonsense numbers, but doesn't seem anyone corrected more than the first 3 mistakes?! And weird unfounded conclusions and framing the previous chat isn't wrong either?I think the entire "debate" is more a low-quality exercise in creative storywriting. Every AI model I've seen during the last 2 years will solve this issue by either reconfirming with the user, or pick one number, or give two estimates. That'll be way more clever than going in circles for 3000 tokens... (Though "reasoning" models do. They'll sometimes have a weird inner debate pretty much like this.)
Regarding the methodology: You really need to include your prompts in the transcript. AI output is always just half of a story. I have a hunch you set them up to fail. First: Seems you prompted for a "debate". And they do mimick a debate. A lot of debates aren't productive, though. Neither in the real world, nor in their training material if it contains online debates. People are stupid, they deliberately lie if it suits their narrative... The AI output reflects it. Furthermore: You start with the "Skeptic". And that's what it does. Be overly skeptical of your initial question and fabricate a different "truth"... That's kind of what a role of a "skeptic" encompasses. And it goes sideways after that. Could very well be your experiment setup that is to blame. I'd say it's in fact likely the cause.
PieFed has a feature pretty much like that. Once I drop a link into the new post form, it takes a while and loads a table with previous posts. Already helped me when I put in a YouTube like I rediscovered. But I had already posted it 3 months earlier... It's a very good feature to have. And probably not that difficult to implement.
I think one of the issues with the AI bubble is, it's going do destroy x trillions of investment money if it does not replace 30% of the human workforce til 203x. I forgot the exact ratio. But people did the numbers on how much people we need to let go (or productivity increase) to make the ludicrous investment bubble in AI work out. So either it happens... Or a lot of money is wasted.
Not sure if money is real or made up, but it's a considerable amount. So if the GDP of the US or China tanks because companies wasted their capital, they might need to let go of people nonetheless, in turn fulfilling parts of the prophecy in a different way... But with hundreds and hundreds of billions already invested per country, the investors are eventually going to need their money back and we're already committed to this kind of future.
This is the first time I heard the term "good AI". Is that a term people use?
But I wonder what Fediverse software she's talking about. I mean we definitely have interoperability. But moving accounts (portability) is kinda notoriously difficult. A lot of platforms don't support it, or don't support it well... It is supported by Mastodon but you really want to read the 2-3 pages of extra information because it's not entirely seamless... It's not the first thing that comes to my mind when talking about why the Fediverse is great.
I really like the overall message, though. Also: always remember we don't have to subscribe to some extremes. It rarely needs to be 100% boycot while not doing anything else to change the bigger picture... Or on the other end of the extremes: 100% demand the government do something while continuing to endulge in doomscrolling sessions on TikTok and X... I think something in the middle would be great. Keep your Instagram account if you like... Use it a bit less, in turn use the Fediverse a bit more... And spend some time molesting your representatives. Tell them it's about time to regulate big tech. Or to get a bit of our personal data back under our control. Would also be nice if one day Jeff Bezos paid more taxes than me... Just saying.
You could try to run Wireguard on a different port which would be otherwise used by some very common service, maybe there's some general exemption for port 21, 22, 53, 80, 443...
Does your bank even allow you to file debit notes, or collect money from other people's credit cards? As far as I know that's never part of a regular consumer bank account...
For payment in advance (if you clients put in the effort to do a bank transfer) I think there's HBCI, FinTS or whatever standards we have integrated in some CRM and eCommerce solutions.
Awesome. Hope this happens. Just make up some clear guidelines on what to expect. Like video, or audio only. Singing, playing instruments, maybe the shittyflute... C64 SID music... maybe include a statement on generative AI.
And better coordinate with other events. Shouldn't clash with Canvas, The Button, Lemmyvision...
Great thing in my eyes would be to attract people who do original content. And write lots of comments/replies. I mean the great thing about Reddit is, how it's not just a RSS reader, but there's plenty people there who reply and post their own art / blogposts / research etc. I think we need way more of that.
Reposting is fine as well, I think? But to compete with Reddit and other platforms, I think we need to do even more.
How does it decide which elements to connect a new "thought" to? I'd imagine that can be quite complex if there's 300 elements on the map and you need to scan them all?! Or does the user do the job of arranging all the "thoughts"?
Yeah, it's a shame, "free-speech" gets abused by bad people and it can't be the scope for any platform. I could imagine it being used for other purposes as well. For example I always wanted to run a Piracy community that doesn't remove links or tutorials on how exactly we do it. But that probably also crosses the lines. ...Some other Robin Hood activity... I mean it could be used for good as well, depending on perspective.
Sry, I don't know if we're talking past each other?! You said you were using Summit or Voyager and I went through the settings of Summit. If you're now(?) using a desktop computer, you'd obviously need to tackle it in some other way. Like the browser extension I mentioned earlier?!
It seems Summit indeed has some feature like this. There's some "Local tracking events". There's "Your Actions" close to the bottom of the menu. If I scroll to the second tab "Completed actions", that pretty much looks like a local copy of every comment and post I sent out. So maybe what you're looking for is already implemented?
I don't have Voyager installed but maybe that's worth searching through the settings as well.
I think it's basically what you said. There's some organizations on Mastodon, some German newspapers and tech magazines, game developers, a few artists and some more known people from hacker spaces. We don't really have that here.