Where is Programming Headed?

A personal timeline of how programming has evolved over the years. Plus a fun 2030 prediction.

2006: The Early Days

  • C++ and gcc/g++ toolchain
  • UI with C++/Qt (commercial license required)
  • SVN for version control
  • Code reviews via change requests and SSH logins for end-to-end review
  • Man pages as primary documentation
  • POSIX queues, mutexes, semaphores, pthreads for multi-threaded systems
  • Deployments done in-person at data centers or NOCs
  • Coding in emacs, vim, NetBeans, or Eclipse

2008: Web 2.0 and Collaboration

  • Stack Overflow launches
  • Web 2.0 explodes; web apps and blogs proliferate
  • Internet becomes the go-to for questions and answers
  • Git adoption begins
  • IRC for real-time communication
  • Specialized roles like SysAdmin and QA emerge
  • Still coding in emacs, vim, NetBeans, or Eclipse
  • Windows ecosystem dominates commercial software development

2012: Open Source and Big Data

  • Open source adoption everywhere; used as a distribution strategy
  • Linux becomes the default server OS
  • Post-financial crash, web startups boom (especially in India)
  • Product Managers run roadmaps, often dubbed the new CTOs
  • Transition from desktop to web application development (still in C++)
  • Data explodes within organizations; Hadoop gains traction
  • JavaScript becomes the default for web UI; Java Swing fades away
  • Man pages replaced by Google searches
  • Open source matures, but Windows dev environments still strong

2014: Data Engineering and Cloud

  • Data engineering emerges; real-time insights in demand
  • Cloud computing becomes ubiquitous; DevOps specialization grows
  • CI/CD adoption increases for seamless deployments
  • Agile and Scrum managers become standard in engineering teams
  • Most development done on personal Linux machines

2018: Containers and Specialization

  • Containers (Kubernetes, Docker) become standard for deployment
  • Mobile apps and frontend frameworks drive specialization (FE, BE, Mobile App dev roles)
  • Full-stack developers become rare; keeping up with changes is challenging
  • Android and Apple stores are key distribution channels; SDK knowledge essential
  • Git is the default version control
  • VS Code, PyCharm, and visual IDEs see massive adoption
  • Blogs and Stack Overflow remain prime resources
  • Open source packages like Redis, MySQL, Postgres, Linux, Kubernetes, Docker, RabbitMQ dominate

2022: LLMs and the Digital Boom

  • LLMs and ChatGPT debut; early versions help with quick answers and code snippets
  • Figma becomes the standard for UI mocks and design
  • JS frameworks consolidate around React; React Native for mobile
  • Firebase and marketing SDK integrations are common
  • Social login is expected everywhere
  • Massive tech hiring post-COVID; bootcamps proliferate as everyone wants to be a developer
  • Developers in AI and ML gravitate toward LLMs and better prompting

2025: The AI Transformation

  • The AI bubble peaks; AI is everywhere and transformational
  • LLMs generate most code; many learn directly from AI-generated code
  • Layoffs are widespread as companies shed post-COVID hiring excess
  • Small orgs adopt AI rapidly, building features that once took teams months
  • Companies question developer headcount and shipping speed
  • Full-stack developer role resurges as LLMs adhere to standards via prompts
  • New grads and laid-off developers struggle to find jobs
  • AI agents that write, test, and deploy code are being experimented with
  • Rise of AI agents with natural language interfaces for specialized tasks
  • Experienced developers with deep systems, architecture, and design knowledge are in demand and leverage LLMs.

2030: Prediction

  • Everyone can now build custom software using natural language and it’s vast majority of software is created and hosted in same environment,
  • Small language models running on local computer and phones busted the 2025 AI bubble as all the data center capex didn’t generate revenue and chinese and singpore AI labs played a pioneer role in doing so,
  • Any and Every software we know of is in a playstore / appstore like store, people pay for store subscription and use as many software as they need. Multiple providers of such stores.
  • Software development isn’t a mass employer like it was once before and robotics is the new sunrise industry at it’s peak but nowhere a mass employer for white collar jobs like software,
  • Every non tech company still has a software development dept hosting developers who run, host, maintain software but build has taken over buy decision for non specialized software.
  • SAAS saw massive consolidation and shutdown leading to lot of unemployment of those who didn’t upskill or merely had soft skills.
  • Most tech teams are now single digit in non tech companies
  • Service industy is of significantly smaller size now compared to before
  • Certain domains still employ and pay top dollars for developers namely kernel development, software for chip design, fintech in HFT, Quant, database engine developer and all sub-fields where knowledge is primarly acquired by getting employed and corporates hold the keys to main branch of the software.
  • Game development stack is completely different as AR is now everywhere and LLMs and world engine has forced them to reivent their stack. Every developer gets a different gaming experience that adjust as they play.