![Cover image for AI and Ethics [Chapter 1]](https://media2.dev.to/dynamic/image/width=1000,height=420,fit=cover,gravity=auto,format=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fpu81893w9idzgziawof2.jpg)
I recently purchased The Oxford Handbook of Ethics of AI. This is a 900 page monster book...with small print and no pictures! (I was kind of hoping...
For further actions, you may consider blocking this person and/or reporting abuse
Is your brain fried after reading that?
Did you actually read it...probably not 😱🤣, but I hope that when you skimmed it that it sparked some thoughts and ideas, and I would love if you share them below.
As I said, I am learning this myself, I am certainly only at the starting line in terms of knowledge in this area, so every idea or thought that people share is valuable to me. 💗
That's some sort of Explain-Like-I'm-Five and before reading to the end, I already found some important aspect that are often overlooked in the current debates, although there has been discussion about AI ethics and responsibility long before chatGPT, see Dr. Joy Buolamwini for example: media.mit.edu/posts/how-i-m-fighti...
To follow up on one point in your previous post about the age of creation is that things are not guaranteed to be of high quality just because they have been produced by human beings. We might even succeed to improve and regulate algorithms and machine learning in a way that the creations would avoid old clichés, not only for the sake of inclusiveness and diversity, but also to avoid boring and uninspiring art and outdated documenation.
I always love ideals, but you could write a whole paper on how attempting to correct for bias is an almost impossible task.
I may tackle it in a future article as I need to give it some deep thought, but the fact that AI does not have temporal reasoning, morals or an understand our intentions means that I think current attempts at correcting biases are lacking and far too "blunt". 💗
I heard that Google has been doing years of research on AI without releasing anything in such a disputable state as chatGPT that Microsoft now wants to connect to Bing search somehow. Probably AI tools will never fully work out of the box without humans learning how to operate them, but then again we can see how far photography has become from a nerdy experimental apparatus to "Siri, take a picture of my dog!"
Maybe it should be up to the end users to make their intentions clear. We don't even manage to communicate with other human people without misunderstanding, how are we supposed to make machines do better? But like you said, it must be up to developers to restrict machines to prevent them from doing harm and spreading misinformation, clichés and hate speech etc. probably both by choosing the training material and by coming up with much better algorithms, as the whole idea of deep learning is based on the idea to take something old and replicate existing information.
Great article @grahamthedev! 😊 If I had to pick a point that stood out to me it’s this one:
“Any attempt to clone ourselves into a digital brain and mechanical presence instantly means the model is outdated. Also the model will never follow the actions our biological self would follow as it will not have the same computational and physical constraints”. I never considered this until I started thinking about AI versions of iconic performers who passed away and how their image is taken from a certain era in their career (e.g., Michael Jackson). The AI might sing and dance like the original, but they don’t have the deceased’s essence. Makes me wonder if the creators of Vocaloid had considered this as they created this AI robot?🤔
Yeah that was one point that I really hadn't considered.
Another related part that really made me stand up and pay attention was the part about "prior computation" coupled with limited computational power, when you put that in the context of human society, is an amazing way of viewing our progress as the human race, as well as that of AI (and possibly the future / limits of AI).
The whole thing is so thought provoking that I am probably going to need days to process and unpack it all. I am glad you enjoyed it! 💗💗
I'm not sure what you mean by the Vocaloid remark. It's just a voice synthesis software, after all. While there has been some machine learning going on in the field (SynthV comes to mind) it still remains voice synthesis software; you put in words and the voice says/sings them.
Well written and summarized, the future as I believe will be with AI. This is inevitable. As I'm working in the technology and business domain for the past years with aot-technologies.com/, I can see that we should start working on creating a new work ethic culture including AI.
Users can easily install apps from Jojoy app without worrying about the security risks commonly associated with third-party app stores.
Another great article filled with GREAT information Thank you! LOVE the "extra fun" you shared. LOVE IT! Drift Boss