This is one of the best unironic ads for self-custody crypto. Stop storing your life’s savings in paper format.

Looks like a wave of the new ios27 Siri is rolling out. Pretty vanilla right now and feels choppy and buggy - however, the responses are really fast.

.@Apple Request for iOS 27: 1. Combine the FaceTime and Phone app. The FaceTime app is useless. Anyone I want to FT I do through the messages app. 2. And for the love of everything holy, figure out a way to fix the phone app. I didn’t realize it could get so convoluted and confusing to make a past calls page. It’s not rocket surgery.

lol wut?

Diet Coke is the single point of failure between the White House and the Mars program. If we run into a shortage we’re all cooked.

Yo @tesla Why am I still on 2026.2.9.1 on the advanced channel for a 2025 MY when my buddy is on 2026.14.x with a 2023 M3

Your car spies on you more than your phone, and it’s not even close. mozillafoundation.org/en/pri…

Here’s how my morning goes: *Starting coding with Claude* *gets into a rhythm and really start to make progress* “You’ve at 90% of your usage for this session”

The State of the AI Address My two cents nobody asked for on the current state of AI. My use case: wide-ranging productivity, vibe-coding, and heavy researching / brainstorming / planning. Anthropic The good: Their Mac and iOS ecosystem plus CLI interface is exactly what I dream of in a platform. Adaptive reasoning is underrated imo. I only wish they’d consolidate to a single model like GPT-5.5 — I don’t want to switch between Sonnet and Opus (and I never touch Haiku.) The bad: As amazing as the infrastructure is, it’s felt hollow underneath. Cowork, Chat, and Projects still confuse me. They won’t always talk to each other and play redundant roles. My advice: Kill Chat, keep Projects, supercharge Cowork. Anthropic is suffering from what OAI had last year: too many models, overlapping functionality, & confusing naming scheme. The ugly: Token usage vs. quality output has gotten atrocious. Model regression is unacceptable. Opus takes far longer to finish tasks compared to Grok or GPT, and the outputs are often comparable. Give me a reason to stay. OpenAI — Put them in the rearview at the start of this year. Used GPT since 2022, so there was a lot of context to leave behind. Claude was innovating at breakneck speed and the jump was easy. The latest Codex and GPT iterations look tempting, but I can’t add a fourth AI to my toolbox. Deep research was the last feature I used before canceling my $20/mo plan. No regrets. Google — Only useful for deep research on the Pro model. Brings a technical and scientific perspective I appreciate, but outside of hard science it keeps giving me inaccurate information. Won’t renew once my premium trial runs out. xAI — Grok carries my daily questions and search. I hardly touch Google search anymore (I use Brave, but same difference.) Grok is fast and clean. Way too agreeable and optimistic though; Opus is more grounded in reality. Grok wins for current news and pulling live information. The image generation obsession is where I lose them. Feels like genuine model, feature, and coding upgrades are on hold so Elon can drop “new and improved” AI anime clips. Not my cup of tea. That said, the Cursor deal and the push for a real Mac app give me genuine hope. Start shipping real advancements.​​​​​​​​​​​​​​​​ Stop messing around. At the end of the day, I value three pillars in an AI. 1. The ability to answer questions fast, accurately, and grounded in realism without bias. 2. The ability to preform extensive deep research on any topic, and get it right every time. 3. Claude level productivity, tool, skill, agent, and coding use. (Claude is so good, the engine powering it just feels outdated right now.)

…annndddd I’ve hit my usage for this session

Claude cowork feels like a superpower across so many different disciplines and projects

DOGE is working on a review focused on strengthening the integrity of disability claims and preventing fraud, while safeguarding access for veterans who rightfully receive benefits. If you are a veteran or service member with relevant experience and would like to help, please send a DM.

Grok and Polymarket just picked my bracket. We’ll see how they did

I’ve needed this for a while! Glad it’s finally here

👀 US AFCENT Chief of Weapons and Tactics, Maj. Claire Randolph talks about the complexity of maintaining OPSEC during June 2025 strikes on Iran, because of "Twitter Randos". "You have twitter feeds of Randos that are just studying where our airplanes go and publishing it..."

🫠