What could a search engine with a chatbot function look like?Back
The big problem with AI (artificial intelligence) is when it presents bad or unreliable information as valuable and appearing 'real'. Google says it will prioritize "presenting facts correctly, ensuring security and removing misinformation". The company is also working on ways to assess whether the technology is being used fairly and ethically. Google has been cautious in recent years regarding releasing new AI products. The company was heavily criticized after firing two leading researchers in "artificial intelligence ethics". The two criticized AI language models, pointing out AI's tendency to reinforce biases in their training data and present false information as fact. Consider, for example, the Supplementary Affair, in which foreign names were excessively checked.
In any case, Google imposes strict restrictions on users' requests to these systems. This was already apparent, for example, on a non-public demo in 2021 of a system similar to ChatGPT. However, how this will ultimately play out in Google Search is still a question to which we can expect the answer very soon.
Other new products in the pipeline include an image generation tool, a product prototyping app, and a set of tools called MakerSuite, that other companies can use to create AI prototypes from within a browser window.