Page 2 - Alignment Problem News Today

A love-hate relationship with end-of-year book lists prompts me to dig deeper into why I read

Ethical AI requires a deep understanding of what there is, what we want, what we think we know, and how intelligence unfolds.

The science-fiction fantasy of machine consciousness seems to be moving towards reality. But what would it mean for humans if artificial intelligence technology became conscious? And how would we know? Computer scientist Brian Christian ponders these complex questions.

Artificial Intelligence (AI) has become ubiquitous in the 21st century, reshaping society, industry, and individual lives. AI algorithms now possess an unprecedented level of predictive and prescriptive capabilities.  What is surprising, is not just how...

Researcher and author Brian Christian shares how collaboration between humans and machine learning systems is evolving. 

And i started reddit out of college, that was 2005. i have built through the social media revolution, ive built through the smartphone revolution, ive built through the broadband revolution. and i really believe what were building through now with a.i. is more significant than all of those. >> on the other hand there have been pretty dire forecasts. jeffrey hinton recently resigned from the a.i. team at google. here he is. >> what we want is some way of making sure that even if theyre smarter th...

Been pretty dire forecasts. jeffrey hinton recently resigned from the a.i. team at google. here he is. >> what we want is some way of making sure that even if theyre smarter than us, theyre going to do things that are beneficial for us. thats called the alignment problem. but we need to try and do that in a world where theres bad actors that want to build robot soldiers to kill people. it seems very hard to me. >> one gentle example, a.i.-generated images of pope francis in a puffer jacket. whic...

Artificial intelligence systems will do what you ask but not necessarily what you meant. The challenge is to make sure they act in line with human’s complex, nuanced values

There are dangers of AI already present but quickly amplifying in both power and prevalence: misinformation, algorithmic bias, surveillance, and intellectual stultification, to name a few.

I blogged previously about AI in a post titled “Immigrants Aren't Coming for Your Job, Robots Are.” So, why am I revisiting this topic now, five years