ChatGPT can now watch your screen to learn how you work

ChatGPT's new 'Computer History' feature watches your screen to help you, and a new report says OpenAI disbanded a key safety team.

Breaking

ChatGPTNew feature watches your screen to learn your workflow.

ChatGPT

2 updates

Expanded summary · ChatGPT

The desktop app for Mac has a new 'Computer History' feature that can watch your screen to learn how you work and help automate tasks.

This optional feature lets ChatGPT see everything on your screen to understand your workflows. It can then suggest ways to do things faster or even finish tasks you started.

For example, it could see a document you're editing and offer to summarize it for a Slack message.

Original reportingChatGPT’s Computer History tracks your clicks and keystrokesRead original ↗

Expanded summary · ChatGPT

(Reported by Financial Times) The company has reportedly disbanded its 'preparedness' team, which was responsible for studying and preventing major AI risks.

This internal team was created to assess if new AI models could pose serious threats and to build safeguards against them.

According to the report, the team was dissolved last month, raising questions about the company's approach to long-term AI safety.

Original reportingOpenAI reportedly disbanded its preparedness teamRead original ↗

Claude

1 update

Expanded summary · Claude

“I think it is fundamentally a crisis of trust,” Amodei said.

Anthropic CEO says AI backlash is ‘fundamentally a crisis of trust’ TechCrunch

Original reportingAnthropic CEO says AI backlash is ‘fundamentally a crisis of trust’ | TechCrunchRead original ↗

Claude Code

1 update

Expanded summary · Claude Code

The inside story of how Claude Code went from an internal CLI to Anthropic's coding agent, told by researchers, engineers and early users who built it.

The Making of Claude Code

Original reportingThe Making of Claude CodeRead original ↗