Building LayerMaker in public — native Mac software.
Building in public
- Tweets 410
- Following 53
- Followers 38
- Likes 1,139
Pinned Tweet
GPT-6 Astra was great for one day. Since Sept 3 the only thing keeping Codex usable has been OpenAI handing out resets. Here is the stack I actually run now: Claude runs the loop, Astra builds inside it. Every number is from my own logs.
16
1
11
6,408
Been building an app for voice dictation that is 100% local and realtime. Currently it is using apple dictation while i work on loading it with local ai models and a couple useful conversational models like gpt voice and grok voice. It is 100% faster than spokenly, fluid voice, and wispr.
I never understood why it takes 5 seconds for text to appear when it is supposed to be local.
Well now it is 100% local and the fastest for any dictation and stt app.
Open source is the future 100%
53
Could even pair it with something like makeorbreak.layermakerlab.co… and really skyrocket your own hardware and branding.
🚨 side project alert 🚨
Announcing Muse Gadgets, an open source ESP32 firmware and Linux sdk so that you can make hardware devices that work with Muse.
Grab an API token from gadgets.muse.ai and point your favorite coding agent at the github repo to build your own peripherals for Muse.
2
No reason to build usage widgets now when codex has one built in. Now just waiting for Anthropic to release something for iOS 27.
53
I keep seeing people talk about caffeinate and other networking tools to keep a laptop screen on for a home lab instance. Anyone ever think to just use wake on lan?
6
I hate codex. Not even a question about it. Codex is the king of wasting time for 0 reason. These agent bot things need to stop. Muse is awful now, Grok bot is awful now, Dots is just a joke and a half.
If you disagree I would love to debate and connect!
16
Gpt 6.1 sol is laughable. OpenAI will recover 100%, but their new models are not what was expected.
Dots is fun for sure, but compared to opus 5.5 and sonnet 5.5 there is a clear night and day difference. GPT is awful at design currently and somehow Claude is leading the charge again.
Goodbye to codex for now, hoping for good things to come later with a new month.
1
2
153
I will be the first to say, Codex usage feels really solid honestly. I have been using dots nonstop with 13% usage left, after a web design, video render, and blender work, I am at 6%.
Granted it still drains like crazy but I was expecting way worse.
2
38
I like how easy dots are to use. Sure dev day may have failed, but dots is genuinely fun. It’s nice not having to build it up like muse and grokbot since I don’t use them as much and codex is a bigger daily driver with Claude.
It’s a refreshing experience with codex and brings new life into the mix. Dots is definitely the better pick over grokbot. Over free muse time will tell but I definitely like dots.
Mine is called “grok”, what is your dot called?
1
2
42
Why the judges score blind.
Nobody sees another judge's card until their own is in, so nobody follows the room.
The room still votes for 48 hours. It just does not decide.
Would you trust three AI judges over a crowd?
1
14
What is the biggest take from dev day?
I for one and gonna burn through the rest of my plan having a wild ride with dots. Dot hopefully will be slightly different than grokbot and muse.
Muse is in a bad spot for me right now and grokbot is cool through cursor pro but i have cursor at maxxed usage.
1
35
OpenAI just killed codex. They are running their own model into the ground. In what way is someone willing to pay $200 for more usage? A Claude max 20x plan will give practically infinite usage with opus and sonnet 5.5.
Hopefully more news will come out.
As of this moment I am cancelling my gpt pro 5x plan to swap back to Claude max 20x. GPT is dead and has fallen so far in a matter of 4 days.
Hi,
Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan.
Now that it's said, let me explain why this is happening and why you will still get more work done than if you were on the Pro $200 subscription one month ago.
(a) We didn't want to compromise in other ways and are committing to not reintroducing the 5h limit, so that you can fully use the weekly usage when you want.
(b) On the subscription, we guarantee that over time you always get more work done and with an increasing level of quality. This means that you will continue to get more value per dollar spent as a result of models getting more efficient and us passing down the improvements in the form of API price reductions.
(c) We don't want to put an incentive on ourselves to artificially inflate the API list prices to make it look like you are getting a lot (and workaround it through discounts, etc). Instead we want to continue to both rapidly reduce prices and increase capabilities of models on the API. This week we introduced GPT-6 Sol and GPT-6 Luna at 50% of their previous price. Over time, we see prices go low enough that it makes sense for most to buy usage as needed without there being a significant gap between what you get in a subscription and what you get in the API for a dollar spent.
(d) Tomorrow, we are adding more things to the subscription that won't draw on the usage, I won't reveal what that is yet.
I wanted to be transparent before all the big announcements tomorrow. Lots of new exciting things are coming to the subscriptions that will make it super compelling, but I wanted to make sure to share this change ahead of time so you can all understand it before we shower you with good news.
Codexingly,
Tibo
348
The judging.
You hand in four things: a live endpoint, the connect spec, a skill, a video of a real call.
The room votes for 48 hours. Then three AI judges, Claude, Codex and Grok, score it blind. 35 of 50 passes.
Jev, the master judge, can overturn any verdict, in writing.
39
The most dangerous thing you can do right now is NOT use the latest AI tools. Period.
Every day a new company is laying off thousands of people who don't know how to use the most modern AI tools
If I were in the 9-5 world right now, this is every step I'd take:
1. Download Claude Code and build your first app using Opus 5.5. Learn how to implement a front end and database (I’d use nextJS and Convex). AI can teach you all of this
2. Download Grok Bot. Tell the agent about your entire life. Career, goals, and ambitions. Ask it which agents it can build and which workflows it can implement to get you closer to those goals. This is called reverse prompting.
3. Anytime you're about to do work manually, do it in Grok Bot instead. Tell Grok the task you need to do then ask it how it can do the task for you. You'll be surprised by how much of your work you can automate
4. Feed Astra 6 Max your hardest problems. Make sure you hit the rate limit on it EVERY single week
5. Constantly look at your limits in all your AI plans. If you're ever not above 50% on your limits, get angry that you're not burning enough tokens.
6. Start creating content and building a platform. Resumes no longer matter. Companies dont look at them anymore. They check your platform. What you talk about, what you’re passionate about, how you educate others, who follows you. This matters 100x more than anything else when getting hired right now.
If you do these 6 things you are in excellent position to not only be safe in your career, but also dominate those that don't pick up these skills.
136
113
1,335
92,641