As AI tools improve to aid software development, it is increasingly noticeable how language-specific semantics and syntax become obsolete, as well as libraries and frameworks. Tools like Github Copilot and Amazon's Code Whisperer translate intent from the developer and deliver the block or algorithm done, with few or no adjustments it is possible to have entire code pages working.
A possible collision and intersection of all languages and frameworks for a single development package is expected for the future, probably as machine-optimized as possible, since we will only interact with AI to develop our systems.
In DevOps we can already find tools like AWS Copilot, a tool capable of automatically creating deployment flows, based on the best configurations made by specialists over the years.
To say that a person has a better capacity to validate an algorithm than an AI is a contentious statement, it would take several years of study for someone to be able to beat a machine that already has all this information, however it is necessary to take some care so that the instructions are provided appropriately.
The AI Developer will be at the forefront of decision-making with the business that is intended to be served, he will have to develop skills much more similar to that of a salesperson than a technician, he will need to understand the business in order to deliver the best solution, he will practically configure the solution intended by the customer.
Obviously, this all sounds simple in words, but those who have already taken the AWS, Azure or GCP certifications know that there are countless possibilities to solve the same problem, and that from the point of view of business, budget and customer needs.
Conclusion:
In my view, highly technical people will continue to work behind AI tools for a long time, as there is a need to code them, until one day they can solve their own limitations and upgrade themselves.
The current developer, who is not working with AI development needs to start thinking more like an architect and salesperson and look to solve real world problems with the help of AI, as the software developer will not exist in a few years, but to do maintenance or migrating existing code bases. Why not create tools to migrate codes or systems?
Despite everything read above, the future looks promising, in all my career I met few people really dedicated to technology, many used it only as a business or a means to an end, from now on, I begin to realize that the market will need extremely technical people (scientists) while the rest will look more like business consultants to meet market needs.
Top comments (0)