Google outlines major AI product updates

[Adam2]

What I find most encouraging is that AI is gradually moving toward assistance that feels contextual rather than generic. The difference between asking a random chatbot a question and having an assistant understand the material you are already working with is enormous

For a student, that might mean asking for an explanation based on the notes they have already written. For a developer, it could mean getting help with code while the surrounding project provides context. For a business, it might mean analysing documents that are already stored in the organisation's systems

The downside is that context can become a trap if the system assumes too much. Sometimes the user wants a completely fresh answer rather than something based on previous assumptions. An assistant that knows too much can occasionally be more annoying than one that knows too little

That suggests another useful feature: the ability to see and control the context being used. Let people say, in effect, use this information but ignore that information. It would make AI feel much less mysterious

Google has the scale to make this kind of interaction normal for millions of people. Whether users see that as liberating or intrusive will probably depend on how much control they feel they have. Convenience is great, but convenience with a visible off switch is better