Oz (@FoxOnTheRunTr) header
Oz (@FoxOnTheRunTr) profile picture

Indie SaaS developer. Digital nomad. I build web & desktop apps πŸŒπŸ’» 1) MacOS File Mng https://xitter.gq/t.co/KTIQUtCq97 2) https://xitter.gq/t.co/2ZKEITy0WT 3) https://xitter.gq/t.co/yM5R5r6oWl

β–€β–„β–€β–„β–€β–„ π™ƒπ™π™Žπ™π™‡π™€ & π™π™‡π™Šπ™’
Joined February 2013
Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)Photo by Oz (@FoxOnTheRunTr)
probability over possibility, evidence over anecdotes. that's what we need to succeed in SaaS, apps, and software services marketing.
4
9
338
"everybody and their grandma are vibe coders now" how to be better then? Automate your taste. AI agents write 8-10x more code but only ships 30% faster due to review bottlenecks. Bring in your real world experience, domain experience, engineering-user connection experience: You need to codify your judgement. Turn repeated bugs , rejected output, better UX/UI into written rules ! or automated evaluation tests, so the AI learns your standard of "good".
4
2
236
Just hit $1,200 ARR πŸŽ‰ after a dip caused by forgetting to refill my AI credits :P Lesson learned: check your email every day for infrastructure alerts or better set up a Hermes/Telegram alert instead.
4
12
334
When a new idea appears, you should ask: Who pays, how do we reach them this week, and what existing work would this replace?
1
6
161
Google says its model development is now on recursive. Let's hope these labs on recursive don't get a stack overflow Carbon looks impressive, will be testing it
1
5
173
As soon as they release it to the public. I think they do it like OpenAI & Claude, with restricted access for government, cybersecurity pros, key companies, etc.
1
48
Okay, I'm convinced OpenAI listens to feedback more than many others. I was just thinking about switching to DeepSeek 4.1 Flash for my daily use since it's a close enough model.
Day 25/ We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today. We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage. And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.
1
124
hey, today was judgment day, the big pruning day. I unfollowed most of the people who don't follow back on X Judgment comes to those who don't play it fair 😜
4
143
my whole day's budget is gone by 9 am. ai is getting expensive, and the worst part: it's slowly becoming a part of me.
1
3
171
Weekend science: testing the output LPF on a 40 m WSPR beacon design.
2
12
281
any of you guys tried to start youtube channels for your apps? with lots of videos to provide value/authority in the niche , meanwhile inserting your own product ads ?
2
11
318
I just heard about a study that found "AI-generated images can make products look cheaper and sell less." I wonder if the same applies to software and UIs πŸ€”
3
7
326
Real world approach to Build in public: Boop won RevenueCat's Shipaton build-in-public award. Its creator handed pet owners printed portraits and used their feedback to shape the app. Building in public gets more useful when people can actually try the thing.
19
18
670
App: "Saved." Me: "Where, though?" A success message should leave you knowing what happened. Show the file name, the folder, or better a button to open it.
5
195
The Secret to Growing on X * based on the latest published open-source X Algorithm.
2
13
530
Detailed May Analysis of X algo source code: xitter.gq/FoxOnTheRunTr/status/2…
X Algorithm 2026 breakdown (just released today) The new open X algorithm code says one thing clearly: Visibility is not about "posting more." It is about predicting who will care, dwell, reply, repost, share, follow you . (and who will mute, block, report, or skip you !) Here's a the playbook if it helps:
16
aka "The X Account Growth Playbook", "What the X Algorithm Looks For", "How to Get Seen on X", "How to Earn More Reach on X" based on my May 2026 post / X Algo
33
Been trying to build an AI fingerprint cleaner for AI-generated images. None of the techniques I've tried have worked. I'm starting to think it's impossible. I can get past SynthID, but not statistical AI checks. Anyone know a good technique or tool to avoid getting flagged by AI detectors for images?
1
4
215
- Mistral launches Voxtral 3: voice agents that listen, call tools and talk back. This type of models when cheap enough, could have lots of use cases - Voxel opens home-robot reservations, targeting deliveries in 2028. I see too many companies, nothing yet hit malls or shops, still a hype? - SnapLogic adds 545 agent-tool templates across 44 enterprise technologies. I like this business, creating and selling AI skills/mcps/tools - Claude designed 17 scorpion-toxin binders that passed lab tests. Antivenom still unproven. I have big hopes for AI for science, yet to see anything tangible - llm-d 1.0 ships: open-source infrastructure to run OSS AI models in production privately. Lots of gov and companies would love to have something like that deployed.
2
5
213
Replacing a customer's spreadsheet means replacing years of little exceptions, workarounds, and habits. A prettier dashboard isn't much reason to switch nowadays I guess. Maybe what's needed now is an agent-first interface.
6
216