Very excited to be working with Erik & the @coinbase team to create the x402 Foundation & bring x402 support to the Cloudflare Agents SDK: https://blog.cloudflare.com/x402/
It's been four months since our acquisition by Casa, and we’re thrilled to announce the latest milestone in our tech integration and market innovation: you can now use a YubiKey to self-custody your Bitcoin.
Our work at Chamber with passkeys and applied cryptography combined with Casa’s unparalleled expertise in Bitcoin security and self-custody made this possible.
More info here: https://blog.casa.io/secure-your-bitcoin-with-yubikey/
It wavered for a brief moment, but crypto is fully partisan in the US gov't. You can find a few exceptions who cross the aisle, just like you can with 2nd amendment supporters. But those are an endangered species.
I had a front-row seat to Adobe's AI work in both product and research, so I felt pretty strongly that it would be quite hard for any startup to compete with the incumbents distribution, data, and budget for compute. I'm still not sure if I was right or wrong.
If you are doing any sort of M&A (or complex fundraising), LLMs are a phenomenal tool. I used chatgpt & gemini to understand nuances of Delaware law & Sanctum + Ollama to help me understand the actual deal docs. It won't replace your counsel, but make you much more in control.
Some fun news from me: Chamber has been acquired by Casa. Nick and Jameson have built an unparalleled security product with deep roots in the bitcoin community. I'm thrilled our product + tech + team has joined them to continue on their mission of enabling everyone to manage their private keys securely.
I'm really curious how MSFT/Inflection made the math work with taxes. If they raised $1.5M and msft pays 650M for a *license*... that 650M is revenue. So call it $500M after taxes. If the investors are keeping their equity in inflection, wouldn't that also be taxable if this is a dividend??
Am I the only one who finds the gap between open source LLMs and proprietary ones to be not just big, but so big as to make the open source model almost unusable from a consumer POV? (different for a builder or someone incorporating that open source model into their own tech stack)
Imagine that tomorrow OpenAI and Anthropic allowed you to run their models locally without any of their current guardrails. What is the realistic worst case scenario that would happen in the subsequent ~6 months?
What is the most private/confidential AI to use? Use case is I want to upload parts of a confidential doc to better analyze and understand it, but want it to be completely private and not used for training data. Gemini if you turn off 'Apps Activity' ? ChatGPT if you are paid?
What is the consensus on the etiquette of tweeting/casting about exceptionally bad interactions with vendors other founders might be tempted to use? I hate complaining, and think a lot of complaints make the individual (not the vendor) look entitled, but also think others should know who to avoid.
