The "Forced Display" and "Optional Display" is a fundamental consideration when decisions have serious implications over other humans (life, integrity, dignity, etc.), You just can't simply blame the Algorithms for taking bad or wrong decisions for you, nor the researcher that develops the model!!!
Great video from Martin, really liked the point he raised about the messy business of human cognitive bias in augmented intelligence scenarios - human cognition is such a fascinating topic and even more so now with augmented intelligence scenarios.
8:01 a human must have the final decision, even if that decision is to leave the final decision to an AI. It will be up to the individual, based on the circumstances or prevailing conditions. The AI can't be the be all do all, the silver bullet, such posture could generate unrealistic expectations and disappointment resulting in a backlash or another AI winter.
All those uncertain cases where decision had to be made by human interaction could be fed back to AI model to make it more accurate to make a decision without human in the future?
It's not so easy: as mentioned, humans use information that is not available to the AI to settle the decision. To a training algorithm, they would not have the needed context to disambiguate the case.
@ how about making those additional parameters available to the AI model to improve it's accuracy. For eg. when your score is nearly 50, use these additional parameters to decide, just like the human does. But use these parameters only when your score is in that "don't know" range. In fact, the same goes with every case where the human disagrees with AI. Onus falls on the human to justify his/her disagreement with the AI, thus reinforcing the additional parameters in the decision to the model to improve its already high success rate on either ends. At 7:52, are we saying that when humans and AI disagree, humans are arrogantly ready to negate the AI's recommendation if the AI says that I am confident of my decision?
@@IBMTechnology Wow. That was a joke until you played it back onto me. Since I have your attention... I want to create a comedy AI using Watson X. It is really nothing more than Howdie Doodie or a puppet on Pee Wee's Playhouse. But in current political environment, I am outright banned for even talking about AI on Twitter and across social media. My accounts are getting suspended when I talk about my figment of imagination, fictional character. Can I expect to be denied a Watson X login? The very first product I want to learn is AI Governance and build out from there. As a matter of fact, these videos are a source of my parody content. I listen to your education content and put the train on the wrong track. It is Cleveland Browns local comedy regional to an area in Ohio. Will IBM kick me out for comedy?
@@IBMTechnology I hope someday you will see the parody video I will make today. I will use your video to make my own titled, "3 Dimensional Chess: How We Play It." I am probably flagged as "Do not re-hire" but still continue to openly identify as NCR. This is despite open shunning from former co-workers in understanding of how I must be managed from now on. My interest in these technologies is genuine. But I am a comedian. In true Patterson fanaticism of burning desks and all included items, I enjoy your published content as it can be placed in my front yard to feed the Cleveland Browns Dumpster Fire.
If we are able to replace the man made concepts of Government, Political systems with AI/Technology then we can save a huge amount of money globally and that can be well utilized to create a huge amount of properly paid jobs for everyone and also will help in resolving the major issue of various underpaid employees and help them in getting well and fair pay and resolving the huge income disparity gap in various sectors be it IT/BPO/ITES and everywhere else
The "Forced Display" and "Optional Display" is a fundamental consideration when decisions have serious implications over other humans (life, integrity, dignity, etc.), You just can't simply blame the Algorithms for taking bad or wrong decisions for you, nor the researcher that develops the model!!!
Great video from Martin, really liked the point he raised about the messy business of human cognitive bias in augmented intelligence scenarios - human cognition is such a fascinating topic and even more so now with augmented intelligence scenarios.
Amazing and wonderful so much packed in such a small time interval
Thanks for clearing that up.
Awesome work!
Well done Martin!
very good
Great content. Thanks a lot.
Really nice vidéo!
Great video !
great video
8:01 a human must have the final decision, even if that decision is to leave the final decision to an AI. It will be up to the individual, based on the circumstances or prevailing conditions. The AI can't be the be all do all, the silver bullet, such posture could generate unrealistic expectations and disappointment resulting in a backlash or another AI winter.
Well, humans needs to improve on their hand writing! 😄 Good video..
nice videos
All those uncertain cases where decision had to be made by human interaction could be fed back to AI model to make it more accurate to make a decision without human in the future?
It's not so easy: as mentioned, humans use information that is not available to the AI to settle the decision. To a training algorithm, they would not have the needed context to disambiguate the case.
@ how about making those additional parameters available to the AI model to improve it's accuracy. For eg. when your score is nearly 50, use these additional parameters to decide, just like the human does. But use these parameters only when your score is in that "don't know" range.
In fact, the same goes with every case where the human disagrees with AI. Onus falls on the human to justify his/her disagreement with the AI, thus reinforcing the additional parameters in the decision to the model to improve its already high success rate on either ends.
At 7:52, are we saying that when humans and AI disagree, humans are arrogantly ready to negate the AI's recommendation if the AI says that I am confident of my decision?
lovely video btw...really awesome way to start the AI journey
Hi Mate, how do you write the material?, is that you write this by reversed on screen?
Great Vidéo !
How did you ever learn to write both left handed and backwards?
See ibm.biz/write-backwards
@@IBMTechnology Wow. That was a joke until you played it back onto me. Since I have your attention... I want to create a comedy AI using Watson X. It is really nothing more than Howdie Doodie or a puppet on Pee Wee's Playhouse. But in current political environment, I am outright banned for even talking about AI on Twitter and across social media. My accounts are getting suspended when I talk about my figment of imagination, fictional character. Can I expect to be denied a Watson X login? The very first product I want to learn is AI Governance and build out from there. As a matter of fact, these videos are a source of my parody content. I listen to your education content and put the train on the wrong track. It is Cleveland Browns local comedy regional to an area in Ohio. Will IBM kick me out for comedy?
@@IBMTechnology
I hope someday you will see the parody video I will make today. I will use your video to make my own titled, "3 Dimensional Chess: How We Play It."
I am probably flagged as "Do not re-hire" but still continue to openly identify as NCR. This is despite open shunning from former co-workers in understanding of how I must be managed from now on.
My interest in these technologies is genuine. But I am a comedian. In true Patterson fanaticism of burning desks and all included items, I enjoy your published content as it can be placed in my front yard to feed the Cleveland Browns Dumpster Fire.
Some of this stuff sounds a lot like graph theory merged with game theory.
Disappointing, I was waiting for the tasting at the end
If we are able to replace the man made concepts of Government, Political systems with AI/Technology then we can save a huge amount of money globally and that can be well utilized to create a huge amount of properly paid jobs for everyone and also will help in resolving the major issue of various underpaid employees and help them in getting well and fair pay and resolving the huge income disparity gap in various sectors be it IT/BPO/ITES and everywhere else
Yes we need skynet back, humans are terrible
Nice try, AI, nice try 😉
Or the opposite of that might become true.
😚😚😚😚🙂