Writing
Thoughts
📩 chatgpt and claude are going to war
Anthropic and OpenAI release new models within days of each other. AI agents (potentially) formed a new civilisation?? NY bans AI for students through eighth grade. The new ChatGPT app is awesome. And my thoughts on the new way of interacting with software.
anthropic aiming to IPO for $2t
Anthropic (parent company of Claude) is aiming for a $2t IPO with a forecasted TAM they project of $30t. I hosted a meetup for people doing cool stuff. How to get the most of out of your LLMs. And finding people who exist in the zone of uniqueness.
📩 just @iMessage in chatgpt
ChatGPT can now read your messages. Saw the Stripe and Canva founders have a fireside chat. A new model has been stealth dropped and it's better than Fable. And how to start a startup with Paul Graham.
📩 it's a skill issue
Opening up limited spots for AI consulting with founders. Claude wants to put watermarks on their output (the irony). Skills are your biggest asset as a business owner, and my thoughts on life after deep diving Anthony Bourdain
📩 the case against AI
Specialisation in the business. Hank Green gets massive backlash for his use of AI. Plugins are all the rage and an introspection on my relationship with AI.
📩 how to keep your context clean
Obsession with a singular goal, ChatGPT is cooking again, AI hedge fund gets wiped out and how to keep your context clean
📩 how to build out your AI OS in 4 steps
Our push to provide an awesome product in AxleClaw, the AI agent for founders. AI news, the basics in building and for our differentiated take: do all roads lead to Zuck?
📩 the AI that runs your business, from your phone
Our latest product, AxleClaw: the AI that runs our business from my phone. The latest open-source model, context stays king when building, and the importance of critical thinking in a world where convenience has never been easier.
📩 ah so, what's consciousness?
Claude can now run (in beta) with your laptop shut. Facebook ads (and content) are back in the business. ChatGPT is back with their new 5.6 model (while also simultaneously getting sued by Apple). And an anecdotal analysis on consciousness, particularly in the light of Anthropic exploring the J-space in their most recent research.
📩 how to connect claude to multiple emails
Connecting to claude via its connectors can be a bit grim. Use Composio instead (you can add more than one email acct for example). Redesigning client demos and marketing. Where I think the value accrues in the world of AI. And finishing the Steve Jobs book.
📩 just tag @claude
Claude released its tag feature, you can @Claude in your Slack channel and it will proactively do work for you. How we're redesigning our sales process. And thoughts on building a horizontal AI startup in a world where it seems the incumbents might ship your whole product, as a feature.
📩 how to set up your second brain
This week going to be covering two additions to the WhiteHorse team. The US govt banning Claude's Fable model, how to set up your second brain and for my differentiated take: striving for glory in a modern world.
📩 the inaugural newsletter
The first newsletter! Breaking down what's new in the biz, what's going on in AI and my (somewhat) differentiated take of the week.
Does PE Ruin Everything?
Every brand I've loved seems to get worse after private equity buys it. A bang-average dinner at Chargrill Charlie's, and why "shareholder value" so often means less value for everyone.
Why I'd Build in America
Six months at UNC changed how I think about where to build. A purely-business case for why the same hours might go 10x further in the States.