Tuesday, January 14, 2020

Liam Ossenfort - Blog Post 1


From the “Walsh 2017 Survey,” the probability of when HLMI (High-Level-Machine-Intelligence) will become relevant is discussed. This survey begins with questioning AI experts, robotics, experts, and outside non-expert; from those who are most familiar with the technology, it is speculated that by 2035 there is a 10 percent HLMI will become relevant, by 2061, there is a 50 percent speculation, and finally by 2109 there will be a 90 percent chance. I am not surprised to see the absence of a 100 percent speculation, for something as complex as AI is very difficult to predict the efficiency in the far future. The basis for this relevance comes to that of ‘when a computer might be able to carry out most human professions at least as well as a typical human.’ These predictions do not seem too far-fetched because of how abundant the use of automized technologies are today.

The survey also prompted the sample groups to predict the likelihood of specific occupations, taking input from both experts and non-experts. The occupations presented were speculated as such: an economist is expected to have a 12 percent chance of being automized judged by experts, and 39 percent according to non-experts. Electrical engineering positions said by experts have a 6 percent chance of becoming automized, and 33 percent brought up by non-experts. A job in technical writing is predicted to have a 31 percent chance of being automized according to experts, and 54 percent from non-expects. The last occupation questioned was regarding civil engineering, experts stating 6 percent likely, and 30 percent announced by non-experts. I found it interesting how the “non-experts” raise more concern for AI taking several different fields of occupations from humans due to automation. I believe this comes from a fear of being underqualified to potentially work alongside these AI technologies, while the “experts” know that the majority of engineering based jobs will be safe, for they will be the primary group to build, maintain, and inform about the technology.

I personally am very excited for the development of AI. Something as powerful as AI has no limit because of how new it still is to us, so it can very much be the new big thing for our generations, similar to the magnitude of the internet.


Comments to Others:

Madeleine Walker-Elders
I personally do a see a future in AI, but I agree with your statement of "I don't think that the robotic revolution is the imminent threat humans should be worried about." You mention "household AI" items often, but i don't think this is where the fear comes from, the real fear comes from the autonomous technologies that will have the potential to make humans irrelevant in the workforce. I agree that household AI is making its presence more and more clear in our lives, but I believe that is a good thing because it trains us to be more accepting of new emerging technologies

Andre Morris
I really enjoy this topic, I have never really put much thought into the concept of a "smart city," but I definitely see the dangers that may come with regards to privacy. I believe certain laws would need to be put into place for something of this magnitude; laws that ensure full privacy and no intrusions into a persons home. Utilizing a "smart city" could increase daily life efficiency and aid in crime prevention, so if executed ethically and under the correct management this is an idea I would support.

Abigail Kryszan
I am always interested in articles following AI. The article you summarized regarding AI learning Quantum Mechanics made me think of the good AI can since there are numerous articles describing the potential harm that AI can bring. Creating an AI that can predict the quantum states of molecules is a huge accomplishment not only in the AI field, but also in the tech field as a whole. This use of AI shows the vastness and flexibility possible within its capabilities, so I believe that AI can be the reason behind many modern-day discoveries and the advancement of technology. Once AI can learn how to design, new technologies will emerge quicker and more efficiently than any human could ever produce.

No comments: