Control Design: Machine builders and system integrators are experimenting with the benefits of artificial intelligence (AI), just like many other industries. The popularity and ubiquitous nature of generative AI has piqued the interest of OEM and integrator engineers, as well. Practical applications of AI for engineers programming code have real benefits and are being implemented into daily workflow.
4 comments:
After reading this article, I was simultaneously reassured and also skeptical in the sense that hearing about the implementation of generative AI moreso as a tool rather than a workhorse gives me a certain amount of reassurance that it isn’t being blindly used, but also skeptical as to how often it is treated solely as a tool and if humans are checking the work of the generative AI to ensure it is properly carrying out its intended function. This duality is primarily motivated by an understanding that for systems mentioned in the article, it is important for innovation that we can begin to rely on AI to free up important time to engage in high level programming for these programmers, but at the same time I remain concerned about how failures in this generation can negatively impact these systems (specifically in edge cases) that then increase the likelihood of harm caused for the system. Overall, if there is enough scrutiny and verification of proper implementation of generative AI then this is fantastic, but only if these qualifications are met.
This isn’t incredible news to me. I know people who use AI in programming courses. I know people who are trying to incorporate this type of AI to be helpful in their day-to-day jobs as programmers. I dont think it's a good idea though. I feel like it takes too much away from our actual programmers and the time spent debugging the AI code could be used towards just writing it yourself. The way I try to use AI with my programming (if it all) and the way they are trying to introduce it in 15-112, is to write the code yourself, and ask AI to help you debug what’s not working. I think this is a fine use of AI. The code that AI writes currently is not great and usually doesn’t work the first time through. But if you tell it what’s wrong, it’s usually pretty good at helping you find the solution.
I find it really interesting that the tech industry is creating this immense hype around generative AI. While I believe that generative AI has its time and place and value, it often feels like it's kind of being shoved in our throats before it's even reached the point where it can do the things it's being told to do. for example, this article is all about how generative AI can help with programming and the backend tasks, but in my own experience, whenever I've tried to use generative AI to help solve some sort of error I was having on Excel or whatever, just to see if I could find an answer that I couldn't find online, the AI was also stumped, or would just gave me information that I was incorrect and didn't work. I think that AI has the capability to be a very helpful tool one day, but there's so much desire to over rely on a product that doesn't really exist yet. In my experience, I think a perfect example of this would be how AI has been integrated into the search features of websites like Google, where you can't escape it and you can't turn it off, but the information the AI provides are often wrong, and any images that the AI search provides don't really help in an image search when you're looking for something that actually exists and the real world.
It seems great, but due to greedy executives it will harm the people that it helps the most. As tools like this progress it fills in previously needed jobs for people that used to take care of the more “trivial” or junior tasks that programmers start off learning early in their career. As fewer jobs go towards more entry level positions it will become increasingly hard to find qualified individuals because there is a smaller pool of early programmers progressing in the field learning what the AI is doing in the background. It also puts an incredible risk for data privacy as information flows through and out of an AI. This data can be exposed at the multitude of transfer points as it flows through the internet back and forth. It also can easily overlook exploits and holes in security that people have the creativity to look out for exposing data in an unforeseen and novel way.
Post a Comment