I, For One, Welcome Our New Self Driving Overlords

15 Replies to “I, For One, Welcome Our New Self Driving Overlords”

1. Hearing her talk of the limitations of AI thinking, it was pretty easy to imagine she was actually talking about progressive liberals. It doesn’t matter the quality of input you provide to them, they cannot reconcile it with their leftist software.

2. The real life example of why a learned talent may not be able to adapt to changing conditions… In the NFL teams have been trying to train Black Quarter-Backs for > 50 years… The closest they have come Is Russel Wilson, but his robotic pass was intercepted to lose the Super Bowl, that proved that the ability to READ & adapt is more important than perfected mechanics…..They always lose the BIG one because they are predictable…

   JMHO

   1. Just this morning … I was wondering out loud … Why are there virtually NO black goalkeepers at the top levels of soccer? It’s a curious fact. So while the EPL are taking knees for BLM … they should actually be insisting that 22% of all goalkeepers be black … right?

   2. How would you explain Warren Moon? Outstanding QB in the CFL, then went on to excel in the NFL. Hall of Famer, too. But yes, one of a very few.

3. 25 years ago I read an annalasis of the quest for AI. Your goal is to get to the moon . You are headed in the correct direction and you are getting closer as you climb a ladder, but you will never get there. It seems not much has changed.

4. AI: just another name for those magical machines that produce fantastical results in some mystical way. Like those expensive yet useless climate and pandemic models (note, they make projections not predictions!), their authors make loads of bread keeping their customers bamboozled.

5. DANGER…Will Robinson!

   1. “Just a moment. Just a moment. I’ve just picked up a fault in the AE-35 unit. It’s going to go 100 percent failure within 72 hours.”

6. People seem to forget that all AI is what we create and we have limitations. There are many who think AI will become sentient if we just do it right, kind of like socialism/communism. I am all for developing and building helpful machines, just make sure there is an easily accessible OFF switch.

7. In the deep recesses of my AI brain … I receive the message: “just ram all the cars that don’t merge properly”. Just smash them to bits … hit the accelerator… HARD! … and you will have successfully merged, and helped all the people behind you merge as well.

   But since I know how the law, and insurance actually work … I sit and grumble out loud. And throw the occasional hand signal.

   1. Kenji, I like the way you think. Lol

8. The danger of TED Talks is that people think that people who give TED Talks are experts in their field.

   Janelle Shane is not an expert in AI; she “…work[s] as a research scientist in Colorado, where [she] make[s] computer-controlled holograms for studying the brain, and other light-steering devices.”

   This TED Talk is the culmination of a <humour blog she’s been running for years about how if you intentionally badly train a Markov-chain-based AI you get hilarious results. This is bordering on a parody of a TED talk.

   1. D R I would say studying the brain is adequate. Lots of things are understandable with out reams of BS and bafflegab. A famous physicist once said that if you could not explain a subject simply you did not know the subject well enough.

      1. Ah, yes. “Anything I don’t understand is easy”, the hallmark of the left side of the Dunning-Kruger graph.

         1. You said that, not me.