Microsoft has been secretly developing its own AI chip, codenamed Athena, since 2019, which it hopes will outperform those currently procured from other vendors, saving both time and money on expensive AI initiatives. The chips are already in use by a select group of Microsoft and OpenAI employees who are testing the technology. In other AI news, OpenAI tested its latest large language model, GPT-4, on its potential for risky behavior and found that the AI was ineffective at phishing attacks, making high-level plans, and covering its tracks on the server. Meanwhile, Google has updated its Google Authenticator app, allowing users to safely back up their one-time codes to their Google Account, increasing both convenience and security.
