There are multiple facts that people totally forgot:
- currently the #LLM models that humanity trained cannot separate fact from fiction, but they are so good at using our cognitive biases to convince us they can.
- training #AI to code, can lead to self improvement of #AI which is an exponential phenomenon that later nobody can stop.
- letting #copilot, or #androidstudiobot have access to your companies' code is a security issue. You should ask before using them.
#io23 #googleio2023 #GoogleIO
- currently the #LLM models that humanity trained cannot separate fact from fiction, but they are so good at using our cognitive biases to convince us they can.
- training #AI to code, can lead to self improvement of #AI which is an exponential phenomenon that later nobody can stop.
- letting #copilot, or #androidstudiobot have access to your companies' code is a security issue. You should ask before using them.
#io23 #googleio2023 #GoogleIO