Anthropic says it is teaching its Claude AI model to control desktop computers based on prompts.
So, yeah… there are still kinks to work out.
It then sends back commands to the computer to keep proceeding with a task.
Anthropic says its Claude chatbot can now control your computer to complete general tasks.Chesnot / Getty
Computer use is general purpose and its unclear what it might ultimately be useful for in practice.
Thus far, AI models perform best and produce fewer hallucinations when theyre provided limited datasets to work with.
Anthropic suggests computer use could perform repetitive tasks or open-ended research.
At the very least it could perhaps be thenew mouse jiggler for Wells Fargo employees.
Anthropic has raisedmore than $7 billionin the past year alone.
Overall, white collar workers still dont seem to be taking up chatbots like ChatGPT or Claude.
So, agents are the new attempt at selling a use case.
The fact that OpenAI itself admitsmost of its active users are studentssort of says it all.
Anthropic in a tweet about the new functionality itselfadmitsthat computer use should be tested with low-risk tasks.
News from the future, delivered to your present.
Two banks say Amazon has paused negotiations on some international data centers.
