We are at the beginning of the AI era. For those of us whose professional career involves developing software – it is an exciting and frightening time. We are facing a modern day John Henry vs. the Machine scenario. Here is my take on AI.
AI is coming for your job if your role…
- is primarily to translate a detail design into code
- requires non-ambiguous requirements
- is to replicate what you did on the last project with only cosmetic changes on a new project
- is being a walking reference for books, datasheets, standards, stack overflow, etc.
- does not require you to know the ‘why’ behind the ‘how’
Or said another way, if your role involves the following you are probably safe at least in the near term …
- creating new solutions for new problems
- analyzing and refining vague problem statements into non-ambiguous requirements
- distilling abstract concepts from concrete solutions
My Personal Beliefs
What I don’t see AI usurping any time soon is the human capacity for:
- creativity
- intuition
- critical thinking
- abstract thinking
Until we can definitively understand human consciousness and intelligence (e.g. are humans just a biological machines with no free will or are we something greater than the sum of our parts) we won’t be able to construct a machine based alternative or reproduction.
That said, AI has potential to be the next big leap forward in humanity’s tool using evolution. And at the micro level this will translate to pain, stress, and economic hardship for those whose livelihoods get supplanted by the new shiny AI tools.
What I tell myself at night to get to sleep.
- As an embedded firmware developer I am in a relatively niche space, i.e. there are other mainstream, lower hanging fruits for corporate greed to go after first.
- I have the luxury of being old and will be long retired before I am replaced by Micro$oft’s Co-pilot.
- AI will never be able to replicate the human consciousness.
Regarding your probably safe list (in reverse order):
– AI can already abstract from specifics. It’s just not very insightful. Time will tell whether its insights get more useful better.
– AI can already turn vague problem statements into requirements. But it’s unclear whether AI can figure out how to ask the right clarifying questions and how much AI can “learn” as it goes through subsequent review loops.
– And as for creating new solutions, AI is already creative. We just label that creativity as “hallucinations,” and the AI developers are actively trying to squash that out of the system.