Keep pulling the thread on Ev.
The US government's ban on Anthropic's Claude Fable model is the first instance of regulating an AI model based on its capabilities.
Salesforce acquired Fin, formerly known as Intercom, for $3.6 billion.
Elon Musk's core strategy involves convincing the market to value his companies based on long-term technological possibilities, which function like long-dated call options.
SpaceX secured a $1.25 billion compute contract with Anthropic and a $700 million contract with Google, creating an AI business with an approximate $24 billion annual revenue run rate.
The revenue run rate from SpaceX's AI business, including contracts with Google, Anthropic, and Cursor, is now larger than the combined revenue from its core SpaceX and Starlink businesses.
The US government's ban on Anthropic's model was triggered after Amazon discovered and reported a method to make the model provide cybersecurity advice.
The US government's ban on Anthropic's model was enacted under the Export Restriction Act, which does not allow for judicial review.
Ev predicts that OpenAI and Google will develop models with capabilities comparable to Anthropic's Mythos and Fable within the next 3 to 6 months.
It is possible to replicate the vulnerability discovery capabilities of Anthropic's Fable model by using open source models with sufficient test-time compute.
Current frontier AI models can autonomously find, chain together multiple vulnerabilities, and orchestrate a cyber attack.
Mistral is reportedly in talks to raise $3 billion at a $20 billion valuation.
Intercom successfully transitioned from a $300 million company growing at 7% to a $400 million company growing at 25% by pivoting its product and business model to AI.