It's not that I think so highly of myself to easily dismiss Elon Musk in this way, as Aegon suggested. I would happily listen to anyone with any kind of experience, even an MSc student if they were trying to do something novel. Or even philosophers who work as part of a team developing AI because they see first hand what works and what doesn't and are normally the ones most widely read of the literature (very useful to have around if like me you're someone more interested in developing than reading). I have had very interesting conversations with people who don't have any formal qualifications in the field. What is important to me is if they have real experience or not in trying to get it to work.
But I have absolutely no time any more for the Elon Musks of this world. Even the media scientists such as Ray Kurzweil and Kevin Warwick who are more interested in promoting themselves as personalities.
The field of AI is still really young despite going 60 years. The pattern is to hype it up and say that human level intelligence is just around the corner and then an AI winter sets in after it's found that the problem is much harder than it looks. Think about this. People were saying this because they were in awe of the kind of processing power you now find in your washing machine.
Normally any success that AI has branches off and becomes a field in its own right. This has led some people to argue that AI will never happen because it's always going to be blue sky research looking for the next success story. Machine learning is the first real commercial success though and what is different this time is that the narrow AI we see now hasn't been called something else (yet). Although it should and probably will get called something other than AI once it becomes an engineering discipline that can be replicated. And Machine Learning only became commercially viable to the extent that it has because of the growth of the computing and the Internet collecting huge amounts of data, and graphics cards giving us the processing power we need. But this kind of AI is very different to what is required for strong AI, or Artificial General Intelligence. But it is a useful tool that the latter will require.
The field of AI has always been one of very few opportunities. There have been loads of really impressive ideas out there which have gone nowhere. Not because they weren't promising or hadn't demonstrated their worth but through lack of funding and opportunity. And once you move away from machine learning that still applies. There are many different subdisciplines in AI that don't get the attention they deserve.
But I have absolutely no time any more for the Elon Musks of this world. Even the media scientists such as Ray Kurzweil and Kevin Warwick who are more interested in promoting themselves as personalities.
The field of AI is still really young despite going 60 years. The pattern is to hype it up and say that human level intelligence is just around the corner and then an AI winter sets in after it's found that the problem is much harder than it looks. Think about this. People were saying this because they were in awe of the kind of processing power you now find in your washing machine.
Normally any success that AI has branches off and becomes a field in its own right. This has led some people to argue that AI will never happen because it's always going to be blue sky research looking for the next success story. Machine learning is the first real commercial success though and what is different this time is that the narrow AI we see now hasn't been called something else (yet). Although it should and probably will get called something other than AI once it becomes an engineering discipline that can be replicated. And Machine Learning only became commercially viable to the extent that it has because of the growth of the computing and the Internet collecting huge amounts of data, and graphics cards giving us the processing power we need. But this kind of AI is very different to what is required for strong AI, or Artificial General Intelligence. But it is a useful tool that the latter will require.
The field of AI has always been one of very few opportunities. There have been loads of really impressive ideas out there which have gone nowhere. Not because they weren't promising or hadn't demonstrated their worth but through lack of funding and opportunity. And once you move away from machine learning that still applies. There are many different subdisciplines in AI that don't get the attention they deserve.
