Knowledge for everyone
We want knowledge about artificial intelligence to be open to everyone. We explain what is happening, how each technology works and what it means for your life, with clear language, familiar examples and stories you’ll want to keep reading.
Follow a curiosity, catch up on something new or make sense of the world. At your own pace. Why we make this site
A system can use a picture alongside a written question, or sound alongside video. Connecting these inputs can answer questions that one input alone cannot, but the model can still miss context or make a confident error.
A recording can be changed so the same spoken words sound as if another person said them. Voice conversion can help with creative work, but a familiar voice in a call is not proof of who is speaking.
A photo service may back up pictures and make them searchable by people, places or objects. Those are separate functions, and the data involved depends on the service and your settings.
A call can sound clearer when software lowers traffic or fan noise around a voice. The system estimates which parts of the recording belong to speech, so it can also remove sounds you wanted to keep.