4 ms·
Speaking as someone doing research in this field, I have an unbelievably hard time imagining this to be the case. The ML community is generally extremely open,
by chillee 7y ago
Speaking as someone doing research in this field, I have an unbelievably hard time imagining this to be the case.
The ML community is generally extremely open, and people know what the other top people are working on. If an AGI was developed in secret, it would have to be without the involvement of the top researchers.
- prepend 7y agoWithout the involvement of who we think are the top researchers. If I were smarter and had more time, I would look for bright young researchers who published early and then stopped, but are still alive.
- mirceal 7y agoyou're assuming that AGI is going to come from ML. While interesting I strongly believe that ML is not going to generate anything close to AGI ever. ML is more like our sense organs that it is to our brain. It can take care o processing some of the input we receive but I don't see it moving past that. super advanced ML + something else will probably be at the root of what could evolve into AGI.
- Kaiyou 7y agoYou probably have a blind spot for people that are not able to speak English and are working in conditions that are kept secret by design. Coincidentally I know someone in that situation who works on AI since at least two decades and who has kept radio silence since one decade on what he's working on exactly.