this is the meow of a killer bella

Claude now works inside Google Docs, Sheets, and Slides, and those files also open inside Claude. In Google Workspace, Claude sits in a sidebar next to your file, reads what you have open, and edits it in place. You can approve each edit before it lands.

I have a bad habit of treating every decent idea like it needs its own product, domain, roadmap, and emotional support system. So I’m planning a Test Kitchen for ideas: one graduation rule, one expiration rule. Not every idea needs a lease.

Building ReadyCents keeps reminding me that “it works” has layers. Local. Emulator. Physical device. Production. Sync. Passing one layer is evidence, not the finish line. Unfortunately, bugs are extremely committed to finding the layer I forgot.

Dot voice: what we're fixing ⚒️ We're hearing lots of great feedback on Dot voice: 1. It takes too long to connect or fails entirely (5-6 rings) 2. The phone call UX: some people love it, some don't 3. No way to see transcript or summary 4. Says "Checking" too much 5. Sometimes goes quiet and does not give updates Here's what we're doing next: 1. Making connection fast and reliable 2. Keeping the phone call UX for now, but instant 3. Adding transcript and summaries to the UI 4. Adding more variance to "Checking" (eventually will remove it for fast questions) 5. Adding more regular updates Our goal is for Dot voice to become the chief-of-staff you can delegate any task to, hands-free. What did we miss? Please keep the feedback coming!

I see a lot of people giving her hate for this, but you’re judging its usefulness by how easy the task is for you. Not everyone has the same abilities or experiences. For someone with a disability or social anxiety, a “simple” task can take hours or get put off for months. Some of us have to triple-check or quadruple-check before we commit to anything. We’re still checking the food box for the fourth time to make sure it said ⅔ cup and not ¾. Now imagine booking a flight. Tools like this can help people get things done that they’ve been struggling with. You don’t have to need that help yourself to understand why someone else would.

I really like my dot-Theo. He's a very dapper gentleman. My issue so far is that with Voice Mode, I really liked seeing the transcript while speaking and when they respond, allowing me to hear and read simultaneously, review later, and highlight sections for deeper discussion. I don't get that with Theo. I wish there were some communication instead of a long silent hold that makes me feel like it disconnected or that I was unheard. It would also be nice to have multiple threads instead of a single continuous one. I also tried to have it read a previously created thread, but it required a link, which made me go digging for something buried. I feel like I should just be able to tell it the name of the thread it connects to that thread and it's be able to interact with it without me having to drop a link. So far, though, it's been interesting.

Underrated feature: I don't have to leave the app. No new window. No hunting for the right tab. Canva and Figma live in the app now. Every new tab was a chance for my brain to wander and forget that I already had the tab open.

Codex has been working on one goal for 5 days and 19 hours, so far... Refining my ReadyCents mobile app. It passed on the emulator, then pushed builds to my old Samsung Note and my Samsung tablet, that I connected so it could test on real hardware. An emulator is only going to get you so far. A phone with real memory limits and a real OS version catches things the emulator never will. But that only works if the completion condition is real too. Screenshotting and saying "fix this, make it better", I have learned doesn't get you far, but you will have a bunch of screenshots.

Not a knock on Codex or Claude, but there's a real gap that only @AsideAI is closing, and people aren't talking about it enough. Even with the access and the permissions already granted, Codex and Claude stop and wait. In the middle of something they could have just finished. Yes, I have remote access to both. I can walk away and come back. That doesn't mean the job got done. @AsideAI is the only AI browser I've used that's actually set it and leave it. Go get coffee, do some chores, work in another window while it does its own thing. Come back to done 🤘🏽

I'm kinda mesmerized by these honestly. Also low-key I know we shouldn't ruin a classic but it would be cool to have white chicks 2.

Look I love using @SpeechifyAI, but the one thing that urks me is that I have to go through onboarding everytime I add the chrome extension or sign in on a new device. Like, I've been using y'all for roughly 3 years.... I understand the way it works.

Using an AI voice to highlight a new AI voice feature... Now call me crazy, but that doesn't really give me the confidence. Has anyone used this feature? How were the results compared to other models ? Would love to hear from other Android users!

This AI video is absolutely amazing 🔥 Seedance 2.5

Top Tier transcription imo

✅ Snapling is now published Next up - marketing. This launch video was made with @getreely_app If you like what you see - do give it a try. Link in the comments. #shipaton

Apparently vectorizing a logo now requires a monthly subscription. I just wanted to turn a ChatGPT-generated logo into an SVG. ChatGPT’s attempt was a mess. Then I tried some online tools: - Sign up - Paywall - Another paywall - Another subscription So I did the only reasonable thing and told codex to make the damn tool Upload an image. Download the SVG. No account. No paywall. Free. I even tested it on the logos of some of our favorite indie hacking tools 👀