Her homepage is: https://bea.cabrerodaniel.com/ Her PhD: work on Crowd simulation - Crowd simulation is not "good enough"
Then she went to Gothenburg to work with autonomous cars. But... ChatGPT came and change her plans
Figure positioning LLMs within Foundational Models. Evething else is AI outer layer: symbolic AI, ML etc.
Erik Knauss said that developers should become more mature not to do something that they need to redo or that will have such a huge negative impact that they cannot scape from. It is a world wild west right now. She wants to created metrics that help assess how good an AI is. She sees a V curve: Requirements Testing
Engineering
People in the middle
fighting with each
other
Perhaps LLM or ML helps create this middle ground Which metrics to use? We need to ask practitioners. The real world is really messy. If the practitioners cannot use the metrics we develop, then why have them?
Nenhum comentário:
Postar um comentário