How will AI transform coding education?
I have developed projects and learning experiences at BloomTech for the past 2 years. Researched, used, and implemented almost every AI tool under the sun…
Here’s how I think AI will transform coding education:
1/ Unit testing is THE skill of the future and must be emphasized first.
As AI becomes more proficient in coding and requires less input from engineers, unit tests will emerge as the most crucial verification method for development teams using LLMs. These tests will validate the output of AI systems, making test-driven development essential. Consequently, developers must excel in creating unit tests to accurately assess AI-generated output.
2/ Like how we teach students to master their calculator in a stats class, we must teach how to master AI in a coding class.
The quality of code generated by AI is directly correlated to the input it receives. Developers should learn to write clear and robust pseudo-code that can be easily understood by AI systems. By teaching developers to break down problems and create solutions in pseudo-code, we can ensure more effective collaboration between developers and AI systems in producing efficient and accurate code.
Similar to a calculator providing incorrect output if an equation is entered improperly, the usefulness of AI depends on the accuracy of the input it receives. To significantly increase developers’ productivity, we must teach them how to properly provide context and problem statements.
3/ Syntax-based evaluation of projects is no longer useful in picking the best engineers.
Developers-in-training should be assessed based on the skills used in their future jobs. In the industry, developers can now readily use various AI tools to find the correct syntax for an operation, making it unnecessary to focus on minor differences in coding syntax. Therefore, school projects should instead be evaluated on creativity, efficiency, and simplicity.
4/ Understanding of product and desired user value will be key to picking the high-performers from the lot.
AI will democratize much of the development process, elevating engineers who truly understand end-user needs above others. By grasping product requirements and meaningful user experiences, engineers can provide better prompts to AI systems, resulting in more valuable output. The top engineers of tomorrow will possess a deep understanding of users and their challenges, leading to improved prompt creation.
What do you think?
Some questions to spark conversation:
- Will AI make us better or worse engineers in the future?
- Should educators embrace or reject the use of AI in class?
- How does AI effect the evaluation criteria of engineers in industry?
