I don't know if anyone else shares my experience, but for me, vibe coding has been a complete miss.
I'm not sure if it's the way I've been prompting, but for anything beyond the banal, LLMs have completely failed to generate what I would consider to be acceptable code.
For anything with WebGL, high-performance networking, or generally anything requiring taste, I think there's still quite a way to go...
Maybe ironically, but it seems to me that we are still _so far_ from being able to generate something like this: https://www.stripe.press/scaling purely from prompting.
> LLMs have completely failed to generate what I would consider to be acceptable code
Totally agree. Recently I asked ChatGPT to write a Z-80 assembler function to multiply two 8-bit numbers. I specified Z-80 assembler to get away from the mainstream. I tried the same prompt three times and got three different responses, two of which were obviously wrong and the third had no obvious flaw but was rather clumsy.
I don't know about others, but personally I feel that the Vision Pro is a solution looking for a problem.
While the 3D / immersive video is surely impressive, and several of the advertised features are truly impressive feats of technology, it feels very much like they will collapse to gimmick-status before long.
I can't picture having this strapped to my face for 8 hours a day. In day-to-day life, I'm just not sure where this fits in... it's like a less-social TV? A laptop where it's awkward to turn around and ask my coworker a question? Generally, the ergonomics of the device just seem... off.
While the device might be good for Apple, I'm not convinced it's good for anyone else.
Super cool, thanks for posting this! I don't know if it's included in the original dataset, but would be really neat if the technical tab had any info on latency / network congestion.
I don't have that info in the current datasets, but I could imagine down the road having a data source with latency information.
Network congestion is another beast. It's unfortunate, because I'm with you that it's really important. Only the network operators can measure congestion directly, and they don't share those metrics. I've wondered if I might be able to roughly guesstimate congestion levels in different areas, but I haven't come up with any terribly satisfying approach for that yet.
It's a bummer. Congestion is super important to the customer experience & super hard to assess short of actually buying and trying out a plan.
Hey! This is cool, I like the idea of a supporting a social network via awards/trophies. Curious how you're thinking of positioning this vs. say, Mastodon or Bluesky? (or Ello, etc.)
Also, if I understand correctly, it's supposed to be a remix of Reddit + Twitter ideas?
Thanks for checking out the alpha. I am a little excited for my first query on this so bear with me.
In terms of positioning this is not meant to compete with any of the existing social platforms but it offers a set of unique features that set it apart from what exists, like distributed governance, constant verification of accounts and content, no ads means no noise and on top of that the way I have envisioned this and the way the roadmap is being worked on, relies on the platform establishing its own global independent network accessible to anyone on the planet. Think of it like a second internet but solely dedicated for (verified) human to human communication. The platform will run two versions, one that runs on regular old internet and a second that stays active if access to the primary network is hindered for any reason (you can hop on either). No other platform offers this nor do they want to because it is an extremely ambitious undertaking.
If I was to sum up the purpose of dygres: Provide access to unlimited communication to anyone on the planet with or without internet access and enable them to make a true living off the platform then hand every verified user voting rights.
Bonus: Add a use case scenario for cryptocurrency.
Per your second query, the MVP is a spartan build because of budget and dev constraints so for now it mimics the functionality of existing platforms for familiarities sake. My approach for this stemmed from my need to understand web development and test my initial thesis of building an MVP and then iterating from there. Future versions of dygres will not look or function the way the alpha does as there will be layers that will allow granular control of the way content is published whether its informative or general entertainment. This granular control will also allow you to create what I think will be the most valuable following on any platform that exists.
I'm not sure if it's the way I've been prompting, but for anything beyond the banal, LLMs have completely failed to generate what I would consider to be acceptable code.
For anything with WebGL, high-performance networking, or generally anything requiring taste, I think there's still quite a way to go...
Maybe ironically, but it seems to me that we are still _so far_ from being able to generate something like this: https://www.stripe.press/scaling purely from prompting.