Upgrade your account to a Community Supporter account and remove most of the site ads.

Reply to thread

Message: <blockquote data-quote="RareBreed" data-source="post: 9088148" data-attributes="member: 6945590">There was a whitepaper on early work for GPT4, and it did an Amazon Technical interview.  It scored 100% and did it in under 4 minutes.  I am pretty sure if an interviewer didn't know they were testing AI, they would hire it in a heart beat.  The hiring process in software development is very broken too, but that's another story.What I find really odd is that almost all my coworkers are taking this like a non-threat and don't seem to be concerned at all.  Maybe because I am the oldest (the next oldest is like 16 years my junior) I've seen more of what corporate America can do.  But despite the Silicon Valley prejudice, I am not an old dog who can't learn new tricks and I have always been reinventing my career.  I have done everything from embedded microcontrollers (before they were called IoT devices), linux device drivers, web front and backend, and data engineering.  At least one good thing that has come of this, is that I find it a very refreshing change of pace relearning the math and understanding how deep learning actually works.  I have always considered myself more of a scientist than engineer, if the old expression "An engineer learns in order to build.  A scientist builds in order to learn" is true.The real question for me is whether AI will replace humans or augment them.  My gut is telling me that corporations will do whatever they can to cut costs.  That is after all, the essence of Capitalism, because if your company doesn't do it, another will (there's some big caveats to AI in particular, namely that training these LLMs is incredibly expensive...so only the richest big companies will be able to do this).  They will "trim the fat" as long as that makes the stockholders happy.  I'm really curious if the managers/bean counters who make those decisions wonder if they can also be replaced?And will regulations work?  As soon as China or some other country starts doing things more cost efficiently, then it will become an AI arms race (in more ways than one).  This is a big reason America has been banning GPU/NPU/TPU specialized processor chips to China in a move to hopefully kneecap them in the AI race.  Of course, this also means that China will be forced to design their own homegrown chips (fortunately, the lithography and other wafer technology processes are dominated by the Dutch and Americans, and it will be a LONG time before the Chinese can catch up...unless they skip right to Quantum Processors or other alternative ML accelerator architectures).</blockquote>

Verification