Navigating the cohesion of digital technologies.
For the past four years, AI has been the hot topic - with each new iteration of products more capable than the last. Industries have rushed to adapt at lightning speed to integrate the new technologies - sometimes a little too soon. Many a Developer and Designer have heard their CEO demand that AI be integrated into whatever they are working on - and have seen the word AI go from a buzzword to a curse word, only barely mumbled under one's breath lest you be lynched by your fellow team mates.
Practically, use of LLMs can reduce the amount of time one must spend on pouring through research documents and forums to find answers, acting as a filter through which we can sieve some of the peripheral detritus that often clouds the research process - turning an 8 hour research task into 4, and effectively improving one's productivity and ability to give concise answers to UX/UI's burning questions.
But where does AI really fit in? Well, first let's clarify, what AI is - and how we are experiencing it. AI as we know it and, is a Large Language Model - developed much like the way that your old Microsoft Word buddy "Clippy" used to be. You prompt the lil guy to do a something, generally selecting from a list of tasks he is capable of doing, and he does his lil Clippy darndest to give you an accurate or helpful response. Okay... It's a lil more complicated than that, but at it's core - ChatGPT (as a well known example) is simply a fancy, expensive and complicated version of that - which gives a much more informed response based on all the information it could gather on the topic from the millions of sources we are training it on, i.e. the internet. This makes it a great tool for research and ideation - because you are able to tap into the collective knowledge of every paper, blog, book or post that is on the internet and condense the findings into digestable, managable pieces - including citations - that can be used to make informed decisions. This gives you the insights of thousands, even millions of educated individuals over the course of history. It is like speaking to the collective knowledge and history of humanity, and I think that is extremely powerful.
Now I realise that we are probably bored to death of hearing about AI, and more likely than not, if you're anything like me, you would run this article through ChatGPT and get it to summarise the points for you. But heres a few ways you can make sure that your research done via an LLM is actually usable for your project:
Ask the right questions.
Phrasing your questions correctly is vitally important. If you ask a badly formatted question, chances are that you will get a poor quality answer. Grammar and spelling arent vital, but be sure to lay out your objectives in a linear way - and be clear about the thing you want it to do. Although the tech is constantly improving, you will do yourself a mental favour by being pro-active (and not lazy) with the way you ask questions.
Check your sources.
This one is particularly for people who are using LLM's to give them answers about something that they themselves are not already knowledgable about. If you aren't an authority on the topic that you are asking about as yet, its a really good idea to check the answers you get before committing to them as fact - use multiple sources and do your own homework in conjunction with the LLM. Remember, ChatGPT also has it's own flaws, and this can crop up in many ways - a majorly problematic one is it's agreeability (it will begin to agree with you if you tell it that it is wrong!). It is not infallible.
Should I worry about AI taking my job?
AI, like any tool that we have developed in our short time on earth, should be used in conjunction with other tools - not in isolation. As human beings, one of our great strengths is our ability to learn, and relearn - and like many of the tools we have invented in the past, it's initial emergence can be scary, and we often shy away from identifying with the things that we are scared of. But I would suggest that wether you have never used AI before, or if you are just getting started - those who understand how to use the tools will always have upper hand over someone who doesn't, or refuses to.
While many industries may shift, integrating one of the many available tools out there - and some jobs become obsolete, this is absolutely no different than how inventions or discoveries have changed the job-sphere in the past. The industrial revolution saw a massive shift in the way people were employed thanks to automation - but new jobs were invented to service that change. And the people who did the best during that time were those who were dynamic enough not to fear the new change, and were positioned to take advantage of the new way of the world. So too must we be dynamic in our adoption of tools that have the potential to help us with any range of tasks.
These days many existing tools, particularly in the digital industry, have already begun integrating AI into their products - from Meta's most annoying, unhelpful LLM, to Google's seamless integration of it into their search engine - "AI" is here, and it's likely not going anywhere anytime soon. You're already using it on your everyday devices, whether you know it (or like it), or not. As a digital creative, it is imperative that we educate ourselves in the best practices of using this new tool, which will ensure our work remains human-centred.
Comments