I killed an obscene number of @OpenArtAI credits bringing this one to life. I learned a lot so it was worth it.
LIGHT HARVEST, a supernatural thriller from Rendered Souls Studios, is now live.
The concept began during @machinecinema’s Director’s Cup GenJam.
#AIShortFilm #MachineCinema #OpenArt
@TokenGremlin Its cute ya'll pinky swearing you are going to jump out of the plane together but I'm willing to bet one of you actually have a parachute.
@stalman I want to be like “wtf are you doing to be so busy that I got to wait for the fourth ring?”
Then I realize oh… I did just give you like… 50ish tasks to work on tho
Well while you are at it, can you make it so that my dots voice is separate from my Chat GPT voice?
And can you also make my Chat GPT voice different from my Codex voice?
Can since you are doing that, can the Work voice also be different from the others?
IN FACT, SINCE YOU ARE LOOKING, WE FINALLY GET CUSTOM VOICES?
As a human living in 2026:
I think we entered the darkest multiverse timeline, and it’s ghetto as hell.
As a Millennial geek living in 2026:
What a great fucking time to be alive.
@embirico@TheRohanVarma@thsottiaux@mxstbr I connected my dot to Discord through MCP. Mentions and replies work, but responses take roughly 1–3 minutes... that's too slow for a natural conversation. We haven’t isolated the bottleneck yet. Is that latency expected, and can it be improved?
As an old millennial who grew up alongside the progression of technology, I’ve watched this movie time and again.
PlayStation vs. Xbox.
Nintendo vs. everybody.
Apple vs. Android.
Now it’s OpenAI vs. Anthropic vs. Grok vs. Google, with plenty of other players in tow.
The giants are fighting viciously for dominance. They really don’t need fandoms turning corporate competition into religious warfare.
Me? Though I’m deeply entrenched in ChatGPT, I actually root for everybody.
Competition is good for us as consumers. The second any one of these tech titans gets too comfortable at the top, innovation slows, complacency creeps in, and WE start getting screwed.
History has shown us this pattern way too many times. Tech is no exception.
Maybe consolidation is inevitable. Maybe eventually they all start eating each other’s faces.
But until then…
May the odds be ever in our favor.
@TokenGremlin I hope my folks in the EU don’t get too bothered by this. Personally, I don’t care if people know what platform I used to draft, edit, or assist with my content.
I’ll pass on the Salem Witch Trials: Anti-AI Simulator.
I’ll be the first to say I’m a Techno-Witch anyway.
@TokenGremlin Well I certainly hope tf not. I have a really nice flow process with Chat > Work > Codex. Three distinct collaborator roles & personas.
Crazy how I got to do employee reduction with my AI staff… AIN’T NOTHING SAFE IN THIS ECONOMY!
(Plz tell me ads will not be on my $100 sub ‼️)
@melvindvivas If you’re stateside and heavily use Codex and Work, don’t sleep on Dots. Last week I was procrastinating with a backlog of 50+ tasks. Now my dot helps manage those projects & tasks; the work queue is actually clearing. The weight off my back is hella unreal.
The way AI has advanced over the past three months has me rethinking all the SaaS subscriptions I’ve amassed. Especially tools built using the same AI platforms I’m already paying for.
Now, before paying another $20/month, I ask Jax (my Codex AI agent), “Can you build this?” If it’s doable, I have him build the damn thing.
Most of the time, I don’t need the whole platform. Just the features that matter to me, tailored to my workflow, my taste, and the little nuances a generic product doesn’t account for.
Not everyone wants to build their own solution. But the more capable AI gets, the more a subscription has to offer beyond what Jax and I can build ourselves.
I didn’t want to be rude, but I’m gonna just say that a strong point is being made here. Mine is fucking great for the most part (though admittedly there have been some kinks). But overall? It has drilled down several weeks worth of work in a day. And had taking a lot of manual work off my plate so I can focus on other things.
And also… all these models have a tendency to reflect back the personality you emit so… oof. 🤷🏾♀️
My #1: let me choose how detailed voice responses are. I get much richer explanations in text, and I want that same depth in voice when I ask for it.
Also, custom voices. I don’t need y’all limiting me to your selections. I want to create my own and assign a different voice to my dot, Codex and Work collaborators. They’re three distinct collaborators, not a set of triplets with the same flat voice.
A few things I’d like to see improved with dots:
- Keep both desktop and laptop authorized so I can choose where a task runs without going back to that machine to grant access again
- An option for dots to launch automatically on startup
- A dedicated mobile/tablet app
- Let me and my dot work in the same Codex or Work chat without lock/user-in-use errors
- Clearly identify messages from my dot as coming from the dot, rather than making them look like I wrote them
- Preserve the intended project and working folder when handing off tasks
- Carry approvals through the full authorized workflow, including checking that it succeeded
- Show whether my dot is working, waiting, blocked or finished, with the last real progress visible
- Give useful connection errors and recovery steps
- A customizable planner and work journal my dot maintains and I oversee