AI has morphed from needing a definition of the acronym to simply being a word. Today, AI is dropped as a keyword similar to tablet or smartphone. A category encompassed. Underneath the acronym AI are the terms Artificial and Intelligence. While PhDs and programmers develop the code and math behind making the models better, every day users of AI want the benefits of a different A; automated autonomous workflows.
The initial rollout of ChatGPT was in November of 2022, and it was a basic writer. Describe what you wanted to create, and it would craft the document, and it would suffice. There was still work that needed to be done to make it more human like, and truly of a caliber that someone would want to take authorship of. Since then ChatGPT and the other LLMs have morphed to be tool workflows far beyond just writing basic texts.
Midjourney released an AI that allowed users to create photos from text and soon after. Google launched Veo that allowed full feature, Hollywood like films. Everyone could be a creator or an editor with just descriptions of scenes and AI would craft the output.
AI is being driven forward today with a use case most pertinent to business efficiency be helping firms write software. Though Claude is the main name around coding, OpenAI and Microsoft had launched CoPilot in 2021 using access to public Github repositories.
Coding went from having to learn languages and libraries and worry about setup, to simply describing a problem and voila, the code was generated. However, like much of AI it is not perfect and requires users to have an understanding of where the output desired is not matching the code output. AI coders have seen liability with setup of lax database, website, or logic issues that highlight the vibe-coded output.
Despite an application not being production ready, the vibe coded apps showcased how people can use their knowledge and desires to create real world output. For software development, the key gatekeeper was a product manager who was trying to balance the output of the team with the features demanded by clients, while adhering to the release schedule promised by business.
Coding AI now allowed not just developers, but everyone involved with the product, to be able to create features in no time. From a more technical GUI design, to walking through the workflow with users to ensure the desired features match the coded output, code was no longer just a developer capability. Add to that, the ability of coding AI to use agents, and now each person had multiple coders at their fingertips who were limited by costs and not time or human fatigue.
What we see with AI now is thinking clearly is elevated. Everyone has the same ability of a model to craft an output. From having agents run daily tasks and make decisions autonomously to writing applications that can constantly monitor, evaluate, and reason the best path forward to automate decisions, AI has transformed the efficiencies of people.
Workflows that would require people to constantly look at information and make mental decisions can now be crafted into processes that are resolved without user intervention. AI is driving a new way forward using autonomous decisions and driving workflow automation. Nothing about its output is Artificial. Long live the new world of AI.