Latest thinking about AI
2022 was definitely the year where AI had its biggest momentum over the past 5 years, at the very least. Keywords related to AI had jumped in interest world wide. ChatGPT, with the chat style AI where users input the prompt through chat command, and the AI reply the chat took the heart of many people.
In this writing, I will share my latest thinking about AI: what’s so fascinating about it, and how it will impact our lives. Feel free to share yours at the comment box.
ChatGPT Genius UX
For me, the ChatGPT prompt and AI reply UX was a genius. It could have built a boring input field continued with a loading UX while the engine craft the answers, and show the result as a whole.
Instead, the ChatGPT UX decided to gave the output where the ChatGPT type the word one by one. It gives experience as if we are talking with a person. For me, this UX improved the perception of the users that the AI was really thinking and crafting the answers thoughtfully.
In the following years (or months) ahead, I believe there will be more humanized AI and robots in the market. Recently, Disney created robot that could connect emotionally with people. Digitally, this will blur the different of real vs fake response. Physically, there will be more friendly robots that are not just to mechanical movement, but also act more humanely, such as making mistakes / creating jokes. This UX will be tricky, yet fun to solve.
Capturing nuance in the Prompt
What’s more interesting was that it can capture nuance of the prompt nicely. The more specific prompt the users can input, the better result ChatGPT can produce. Its ability to understand sentence was also impressive. It can differentiate subject, predicate, verbs, and produce different results based on the different inputted sentences.
What this means for us, in my opinion, is that critical and detailed thinking is become more and more crucial skill to stay relevant in the job market. In the long run, what will differentiate us from others is that the ability to produce high quality prompt, as well as to judge and improve the quality of the ouput.
What’s produced by ChatGPT so far are theories; to make sure it’s relevant as much as possible, the users will still need contextualized the output. What this means, at this stage, I believe AI still can’t replace any role perfectly. Instead, it will be complementary where it will speed up our works.
For product managers, AI can easily produce detailed PRD with user stories faster than any human being can do, given the same prompt and context.
What I’m more interested for the next phase of AI is, on how the AI can improve the prompt details quality. So, instead of the AI just receive the prompt inputted by the users, it can also guide its users by asking more detailed questions so it can produce a better output.
Open API as the acceleration of AI
One of the most expensive cost to build AI is the amount of data input it needs to train itself. To achieve the level of what ChatGPT able to do, it requires years and billions of data trained.
With ChatGPT provide its open API to the public, I believe there will be more AI companies launched. These companies will contextualized the AI based on certain task to be done. For example, description optimization, image editor, music producers, etc.
Bigger companies will also in a race to implement AI in its companies to stay relevant. FOMO will be the biggest driver, as they don’t want to be disrupted by the upcoming AI wave technology. Shopify, for example, has introduced Shopify magic, a tool to produced description and title easier for its sellers.
The main challenges for these players remains on how to create differentiation compared with other players. I believe, the easiness of the inputted prompt, as well as the quality of its output will be crucially important for these AI companies.
What’s next?
What’s crucial for AI is the data training availability and how the data being processed. Personally, I am soo curious on how Google will play along in this field. With the ChatGPT launch and its collaboration with Bing, Google set red code in its company. With one of the biggest (if not, the biggest) search engine in the world, it surely stores a lot of data. If Google can execute this well, I believe the next wave of smarter and even more mind blowing AI will be born.
