Skip to main content

we went to see stardew valley symphony today in newark, nj. it’s a game that me and my wife put ten years into, made virtually millions and bonded over it, 3000+ millions apart — it’s amazing how pixels brings people together, remembering a lot about our interactions while listening! so good, a must

now that i’m able to work i’m bringing back and refreshing all my additional gig worker areas, put a domain on ugc50.com yesterday and i’m going through and tidying that up, making it responsive seo and all that, will need an astro refresh for pSEO but will do for now, next is the carrd site

I finally got on the new sora2 and I have to admit it’s pretty darn terrifying, this is something new in the hands of the many I just don’t know where it’s going to take us — for now I’m just using it as another augmented audience redirect and to live out old memories of the before the burn times.

the price of rent elsewhere is much better than here, problem is the amount of people that voted a way we don’t vibe with increased, so it’s either move there and live on the beach or keep slinging pizza, bagel and martial arts around here for a short while. my therapy is gonna need therapy. lulz.

found this interesting that a service called recurrentauto.com has this really nice travel range visualization for an ev owners, we are absolutely gorging on the remixing of data in this day and age – they really do make so many intrigue datapoints. i’m a sucker for them, guess they know that

nasty bug in my scheduler, hopefully fixed now, added some locking, also you can use cursor keys now to go through “blips” on sett.ee and click image for modal full size, i really wanna get back to working on moving render code to prod and sidebar so be nicer/friendly – also slop feed soon!

here we go into the breach my friends. october just showed up like that, september flew, end of the year and mad dash christmas shopping will be next, until then, fire up the gpus and let’s render something different. trying to find my feet again, spent few hours debugging that lag, such is life.

just a headsup bartender6 on macOS Tahoe is a mess. remove it, had my WindowServer lagging like a mofo.

so leonardoai/lucid-origin is an off the chain image maker and sora2 just dropped an app with a ai slop feed but somehow it’s giving off back in the day twitter just launched vibes, also suno music has a feed thing launching, all these infinite noise machines, we will pay just for the quiet soon.

i wonder if when the aliens do arrive (ahem, it’s a comet right) they take one look at what’s going on here and think we are just stuck in some kind of recursive fever dream, take a data dump of the now and just head on out. i mean it would make a lot more sense.

openai.com/index/sor… looks actually insane from the videos i’ve watched, every day i see another model coming out, i’m trying to categorize the ones i want to use as part of my process and flow but it’s changing all the time, some are better than others but it’s gonna come down to cost.

probably going to rebuild niche2025baby.carrd.co in ASTRO so i can do a bunch of SEO/pSEO things ( hard on carrd ) because the market for carrd templates has grown and i really wanna have that stuff in a strong position moving forward, i hope they bring out ai features next year. onwards!

.. .hitting the ground running, monday was a bit of a bust but we are straight into fresh work on my usual UGC channels, feels good to be back, little different because it’s a year later and i’ve got to do a lot of adapting and refreshing, new lick of digital paint if you will, also, AI chat, oooo!

been enjoying having a decent camera system on my phone again, this was hotpot recently, pretty rad spot, great food, little pricy, but was good to do it all again similar to our time in michigan. gotta think of places for our anniversary now, drawing a blank atm to make it special.

my adhd is on full tilt today, i’m having to write things down just to keep myself in alignment, next i’ve gotta sort that in terms of what’s gonna give good results quickly and using my time sparingly, just getting over the hump of decision paralysis can be quite the quest on a monday.

don’t think it’s fair that as a 2025 genX human i’ve not dedicated enough time to me finding the correct clothing for my roller skating weekend dreams. now that i’m able to work again i expect i’ll be able to accelerate acquiring skates and shiny tops with killer glasses in no time. wait this space!

i'm building an rss scheduler with ai and so far so good

well good morning.

i’ve been steadily been building out an rss scheduler in recent weeks, maybe a few hours (or more) a day back and forth, using agents for different sections, going over security, optimizations, fixes, ui/ux and something super fast and clean that i can run on my local machine that talks to api’s and does what i need it to do – this is just the first version but i’m pretty happy about the progress i’ve made with it, it’s not production ready enough to package it up and sell it to others but it’s getting some heavy testing!

some of the features are as follows

  • support multiple platforms, each with their specific rate limiting/requirements
  • check that i have local connectivity for when my connection is off
  • what platforms are up and talking, checks for the api and general sites
  • what happened recently and when the connectivity was restored
  • posting windows! (when the best time of day on those platforms is best)
  • when the next post is going out
  • a “slots” scheduler so they get evenly padded/gapped out so not spammy
  • max number of slots per day with automatic rollover for next day posting
  • natural language analytics and at a glance
  • pushover integration for when scheduler is up/down/posted/added to que
  • quotes for inspiration each day

i’m probably leaving out crucial details of all the features but when i do finally push it i’ll probably end up putting it as a free version on github and a packaged up version for people using a mac with a nice app, tray bar notifications and all that good stuff, right now i’ve got too much on my plate to do that but i expect in time i’ll get into a flow and start putting out time saver apps like this for content creators and people who stream/game on the web – so many ideas to build out!

my plan is to build out a decent framework to rapidly prototype ideas and the websites around them and give away free versions of the codebase for people to adapt and build on themselves with ai and to leave a donation.

i’m thinking scale here, stuff that can be easily maintained, easily regularly promoted (think seasonal, think dynamic, think scheduled calendar ai marketing activities) and just a creative mental flow – idea, deployment, marketing, donation, next one.

i really wanted to see how far i could get with ai building a smart scheduler that i ran locally and then eventually deployed to a VPS somewhere in the cloud – obviously i’ve got plenty of testing yet to do but i’m super happy with it so far and i’m learning so much in the process. i might not totally understand at the code level all the things it’s doing but i do understand best practice and i believe we have a good human vs computer thing going back and forth.

what are you building out there?

craftgpt, the stuff they build in minecraft!

YouTube Thumbnail

i’m not even sure how to mentally process this.

every morning i think to myself my list of things i need to at the bare minimum to do compared to the stuff i’d prefer to be working on grows longer and then you see some guy just on youtube building a type of chatgpt inside of minecraft. .. and. .. it kinda works?

now of course he’s obviously using a tool to create all the attributes inside of minecraft (it’s not a by hand thing) but still, the fact that he got this up and running is kinda mindblowing.

don’t expect it to overtake your claude code or fav model but props to the guy to even pulling this off.

wow.

working in the cloud/remote/in-office is a total crab shoot these days, i’d rather wake up and have ai tell me that i have problems with the machine, software, dns, whole bunch of bs before i actually get to the work bit and it’s 10:30am and i’ve yet to hit the coffee. still, got the green card now!

. .. busy learning about pSEO, might have to also build my ideas out on astro and import my current posts in from micro.blog, but still stay here of course as i like the platform and the tooling, also improving fiverr ads, seo, gigs and aura maxing the heck out of my offerings, also agency setup!

running on impulse power today, a week of poor sleeps have caught up with me. low grade flu, runny nose really effects already bad sleeper. shiz happens i guess. trying to take it easier on myself today, shower and sun time required i think to break up the day. onwards brave warriors.

ugc creators need to evolve with the times

ugc creators that do not offer ai services moving forward will get left behind, not because their authentic services are not required it’s just that brands are going to change too.

they are going to need not just a ugc creator but ideally a one-stop shop that can also advise and bring their own style and creativity to the job at hand.

the playing field for creativity is limitless now, it’s how you deploy that’s going to be the difference between getting a repeat job or not.

currently i’m looking to use cloud services, replicate, fal, render for in the cloud processing of my ai creative ideas, a bunch will be failover or for platforms that managed to roll out new models faster than the others.

i’ve already built an rss aggregator and sharing to social tools que manager using vibe coding and i’m looking to improve on that and ultimately sell that as a standalone product.

i’m thinking both desktop/mobile in the build out for the needs of a ugc/brand advocate influencer that’s on the ground — to put tools into the hands on the ground and for the backroom team handling brand needs remotely for in time relatively live needs.

now that media is blending all together it makes sense that eventually your video editor, stream manager, media management will at least be done by a bunch of ai tools — how you utilize and deploy those things, first, fast, timely are what brands are looking for.

but also i see a near future place for ugc creators to share the behind the scenes of the creative process and the TTL (time to live) from accepting the order and packaging the results for the client.

eventually, a lot of that creative process can be automated (especially with seeds) or at least brand kit ideas can be built out faster.

i don’t think a ugc creator/influencer will be just “waiting” for jobs now we have these tools, literally any tool can be built.

creative ideas can be put together and time saving elements can be built between brand and creator, the ones who make these time saving tools will also increase their audience indirectly.

after all, speed wins.

working on pulling together all the models i’m currently using for images, video and audio, gonna rank them for usages, need to make a 1-pass blast at each model for review/preview, select and then more features, also failover support for other api’s and services, fund manager etc, gonna be useful.

i always turn up to the keyboard in the morning like this (mentally, not physically) ready for action, i have lists on lists and lots of things to deploy and roll out, some take longer than others, but today i feel is going to be an aggregation and split apart day in bear - thursday came around fast

the areas i’m focusing on next

back to udemy, fiverr and selling niche scripts/apps with a side order of flipping to the mix is my current path, potentially going to re-train for new roles that are currently appearing because of ai.

oh and i need to learn to drive all over again, so that’s nice. at least, i need to learn new measurements or do the calculations and i need to be able to pass tests that seem to be more vindictive than useful, but hey ho, you put in the coin and you take a spin.

with fiverr i’m going to be adding a layer of ai spice to the mix, a middleman agency/studio setup for people who want someone to handle their ai workflows but utilizing me as the person they hand things off too.

concerning udemy it’s been good to me over the years and i know there is a lot of training and teaching to be done, specific areas of society using ai tools to enhance what they are already doing or to break out of where they are in life and create a better life (myself included here)

i can visualize it very clearly for both udemy/fiverr as i’ve done those both at different times over the years and building for both has’nt really changed — fiverr might be a bit more involved because i really need to keep up on pricing on the saas side for the rendering.

i think i could build a middleware tool for me as a creator relatively straight forward in terms of actions, the idea would be something more using comfy.org and nodes but i’m not there yet — another part could be n8n once i’ve built those other elements.

obviously pricing, bandwidth, failovers, remaking and such like is going to be a fine balance to start with so i’m building in some wiggle room because i know what fiverr clients tend to be like.

of course a lot of this depends on the “clouds” and they get costly, have to balance out what’s worth the time, where i can save time, where i can do stuff while mobile to kick things off and then also deploying a homelab that can maybe take a load of (costs) and do ai inference locally across the network.

i think we both as remote workers are heading that way at the moment, cutting costs, working from home, having all our assets locally, doing ai experiments locally, backups, storage, location security — kinda where my head goes when i’m mapping that out.

also i have to take in account that i’ve been out of the loop of the “demand” side in the last year. it’s like i took a gap year and i’m now re-entering the job market when i know that apple is out here redesigning chipsets in 6 months — a lot changes faster these days.

might not be an audience for the stuff i make on fiverr anymore as a 50 something year old man, that market last year was crazy as the market had enough of twenty something inflencers and they were just not doing well in a/b split tests in facebook metrics.

i still think my “age” of ugc could do ok, i’m sure they just want the tried and tested, real to camera, reletable stuff but i’m hoping i can build an ai agency component on the side of that too. i’m not truly sure what that looks like at the moment for brands i work with as they already have a lot of that in place but costs/speed/tech could make that unruly.

the content creator in the ai age is really appealing to me.

i’m seeing a bunch of decade or old creators that are finally using ai to make thumbnails or re-write seo ready titles and descriptions, even conversation starters in the comments, i’m sure they will expand on that.

i can certainly see a world where every part of the conversation tree is remixable, including down the website, the original video shot, b-roll and just every interation possible.

not sure where reputation management sits now but i’ve really not heard a lot about that these days although being here, i’m more interesting in the parody laws! :)