Keep pulling the thread on Muse Spark.
Anthropic unveiled a new AI model named Mythos but is withholding it from public release due to its advanced hacking capabilities.
Meta has debuted Muse Spark, the first model from its superintelligence labs led by Alex Wang.
Anthropic's Mythos model discovered thousands of zero-day vulnerabilities, some of which were quite old.
Anthropic's Mythos model operates agentically, autonomously scanning codebases to find vulnerabilities without requiring human steering.
Two days after being acquired by MyFitnessPal for approximately $100 million, CalAI was breached, resulting in the theft of 3.2 million user records.
Jason predicts a transitional period where cybersecurity will worsen significantly as AI tools make it possible for bad actors to instantly hack any website.
Anthropic reached a $30 billion valuation within five years of its founding.
Amazon's Trainium chip business is now a $20 billion annualized business experiencing triple-digit growth, according to CEO Andy Jassy.
Amazon's $20 billion internal Trainium business represents revenue that did not go to NVIDIA, equating to roughly 10% of NVIDIA's revenue.
Jason believes that most incumbent software companies are only building AI features that are about 60% as good as the best-in-class standalone solutions.
Jason argues that incumbent companies cannot successfully monetize AI features that are only 60% as good as standalone products, as they will have to be offered for free.
Jason asserts that incumbent software companies whose AI agents are only 60% as good as competitors are in a "slow death spiral."