Mckay Wrigley

i love the united states of america. a recent favorite example of why i love this country so much is this post from peter steinberger. it’s so representative of everything that makes america great. “ye are the light of the world. a city that is set on a hill cannot be hid.” at our best, we are a bastion of freedom and opportunity for the entire world, and we should continue to both celebrate and protect the values and principles that make the usa and the american dream so wonderfully unique and possible. i live in a part of the country that’s particularly patriotic (if you’ve never seen how we do 4th of july here in utah, you should!), and today on the internet i felt a lot of that same pride that i see often in my community. thank you to all of my fellow citizens - and our friends around the world - who spoke up today. and keep speaking. this is only going to escalate as agi approaches. but today i’m very proud to be an american, and i’ll always continue to be long the usa. god bless the united states of america.

anthropic. the life, liberty, and pursuit of happiness company. ai allows us to raise our ambitions. but as the technology advances, as a country we should never lower our standards. let freedom ring. 🇺🇸 🇺🇸 🇺🇸

as a very, very proud american citizen i stand with dario amodei and appreciate his willingness to defend the civil liberties of other american citizens 🇺🇸

Anthropic - here’s what we want… You *can* do this: <list> You *can’t* do this: <list> And then keep the messaging consistent between docs, tos, and what team says online. Would solve this issue immediately.

anthropic undersold opus 4.6. its ability to infer intent feels like the biggest upgrade. genuine world-class coworker. hard to not call these things agi…

opus 4.6 with new “swarm” mode vs. opus 4.6 without it. 2.5x faster + done better. swarms work! and multi-agent tmux view is *genius*. insane claude code update.

Re quite literally

Re @nikitabier ah here comes another wave. all clearly real users and definitely not easily identifiable bots.

people haven’t properly priced in codex + cerebras. i would *easily* pay 5x for 2x speed. and it will be much, much faster.

claude code's new task system is soooo good. total game changer for multi-agent work. agent swarms that *actually* work are coming soon.

nobody ever listens

me after hitting "approve" for the 9,646th time in claude code without looking at a single line of code it wrote

everyone started talking about an ai bubble when they actually weren’t hyping things even close to enough.

claude is ozempic for knowledge work

fyi: you can build your own claude cowork style apps with claude agent sdk! i’ll be open sourcing my app this week. i’ve been hacking on it for a while, and i think it’s actually really good! this app format will be the ai app-layer trend of the year. watch for a 4min demo.

claude code + opus 4.5 injected the immaculate hacker vibes back into ai that we haven’t had since gpt-4. everything is new + fun + weird again. you can feel it. another oom of new ideas & latent economic value is waiting to be unlocked. and building has never been this fun.

the claude agent sdk will do for knowledge work in 2026 what claude code did for coding in 2025. want to live in the future? build your own agents with it. you’ll have a personal magic tool belt that the public won’t have for another 6-12mo. it 100% feels like personal agi.

RT david friedberg Re why not just raise income tax rates? because your real intent is not to just “provide healthcare”. you’re masking that you are proposing the creation of, for the first time in the 250 years of this American republic, an organized government seizure of private property from citizens. you’re calling it a “wealth tax” or a “billionaires tax” or “millionaires tax” or whatever nom du jour polls well. but at the end of the day, it’s the seizure of private property from citizens by the government. citizens that earned money, paid their fair taxes on those earnings (53% if they live in California) and are now being told they need to hand over after-tax assets because the government has failed to provide promised services with the revenue it’s collected, and are now re-casting their own failure to be a socio-economic inequity that must be justly resolved... a slippery slope that has never gone anywhere good (see economic effects in USSR, Cuba, Venezuela, France and Norway wealth tax etc.) the American founders fled tyranny in Europe and this amazing nation was populated by immigrants (myself and your parents) from around the world not just looking for a “better life” but for a place where they could have freedom from tyrannical governments that can take what they want from private citizens. a great nation borne of property rights, the rule of law, and endowed freedoms to believe, speak, or act. these principles led to the greatest run of innovations, successes, and widespread increase in prosperity, for all citizens, ever seen. the citizens, the individuals, not the institutions, delivered this progress. those who invented, who toiled, who bled, who sacrificed, who took risk and persevered, who led, and who changed the world, are not charlatans, kleptocrats, or oligarchs. they’re what made us all better off. prosperity is a measure of america’s success, not its failure. it is your principle that is so offensive, as evidenced by the broad dis...

Re literally anyone i know who is really good at factorio is immediately top 1% of agentic coders

1/5 The more I code with Opus 4.5, the more I think we’re 6-12mo away from solving software. The model is pretty much there. I’ll build like 3 versions of an app in a few hours just to explore options that each would’ve taken me 1-2 weeks <1 year ago. It’s getting weird.2/5 As AGI gets closer, I expect things to start feeling increasingly weird. And the weirdness is definitely starting to accelerate.3/5 Opus 4.5: “Would you like me to do A or B here?” Me: “Actually can you create ephemeral test pages for both that are interactive presentations of what those solutions would look like fully implemented and explain the why, pros/cons, etc? And do C and D too just so I can see.”4/5 People underrate how massive the network effects of Claude Skills alone will be in driving performance over the next few months. Anthropic gave a great talk at AIE about how skills are kinda like a form of continual learning and I agree. “don’t build agents build skills” on YT5/5 Strong agree https://x.com/paultoo/status/1999245292294803914?s=20

Had a blast with @ryancarson and @thorstenball this morning! Fun chat about agentic coding and Opus 4.5.

It’s not AGI until >90% of OpenAI, Anthropic, DeepMind, and xAI employees stop contributing to their 401k.