It’s a little over two years this month since OpenAI launched ChatGPT publicly. I suspect many amongst us went through the stages of curiosity, amazement/incredulity, awe/fear, to perhaps being spurred into action to figure out how best to stay abreast and not become irrelevant fast. I confess that at the beginning of 2024, we at Nayamode started taking AI more seriously as a genuine disruptive threat to our work on the creative side of things. Since then, we’re proud to have launched some of our own AI-powered solutions – from videos, AI-generated content, chatbots to significantly enhanced events.
Fast forward to today and the picture looks different. From reports of AI training running into a wall to the debate over whether there’ll ever by real ROI from the mega-investments in AI, to whether AI is actually delivering on its promise to make life so much better for users, as evidenced by apparent struggles for Microsoft Co-pilot. If you’re even a casual observer of the space, you may have noticed a marked difference in the posts related to AI from early 2024 to now. Where before it was all rah-rah about all the amazing things AI could do, it’s now turned more to discussing how to facilitate adoption/deployment and derive value from the technology.
Here’s our take on it.
It’s going to be all about productivity for now. Creators, you can rest easy….at least for a bit.
AI, or Generative AI as we see it today is great for several tasks and seems to struggle at others. Specifically, tasks that require consuming vast quantities of data and creating intelligence or useful data from, are well performed by AI. In this category I’d club activities such as transcribing meetings, providing summaries and notes/action items. At Nayamode, we take it a couple of steps further. Where we are able to unleash the human mind’s ability to ideate while speaking vs while writing. What in the past would have taken hours to do if you were say, writing a requirements doc for a new feature or product, now you can free-speak into a microphone, have the recording transcribed, summarized and converted into a Business Requirements Doc (BRD) in literally a few minutes. You can go further and have AI generate a functional spec from the BRD with specific technology constraints as you deem fit.
Similarly, we’re getting great results using AI to enhance our ability to get answers on questions ranging from HR issues, business policies or regulations in different states, light legal advice and so on. Stuff that exists on the internet, has been scraped/digested by the LLMs is made way easier to find, consume and use.
On the other end of the spectrum, every time we’ve tried to use AI to do anything from help brainstorm ideas, generate suggestions for brand names or taglines, or even generate imagery that goes beyond the very basic, we’ve run into significant issues. The responses are banal at best, largely uninspiring and would make any middle-schooler look like a copywriting genius. And they can get worse, as you try to fine-tune your prompt to get to the right answer or response and AI keeps misunderstanding what you really need, makes things increasingly convoluted and at some point, quite far removed from what you wanted from it in the first place. Leaving a very frustrated user at the end.
We’re sure this will change as capabilities evolve, but at the moment it does feel like as with any new technology, AI is finding use-cases that it is well-suited to, while clearly flailing at others that are a stretch. Our suggestion is to harness its power for the stuff it’s really good at (and there’s many of those) to dramatically reduce the time you spend on mundane, repetitive or well-defined tasks.
But if you want to come up with a great new brand name, tagline or logo? Well, it’s a good thing humans with natural intelligence are still around to help!