Advice to CS Students in 2026
Six pieces of advice for navigating a degree in the AI era, each with a pattern to adopt, an anti-pattern to avoid, and the inner statement that gives the anti-pattern away. Available in English and Hebrew.
Guides for Students
Everything a student in one of my courses should read before starting, and keep open while working. How to approach the degree, which assumptions to drop, what a strong project has to demonstrate, and where to find worked examples, course assistants, and the textbooks behind the courses.
Start Here
Read these two before choosing a project. They shape how everything else is judged.
Common Misconceptions
Ten beliefs that quietly limit what students get out of a project-based course, from mistaking a working demo for engineering skill to treating a course as content to consume. Each comes with the sentence that gives it away.
Project Standards
What the project has to demonstrate, at two levels of detail. Use them as a checklist while the work is still taking shape.
Project Requirements
Twelve requirements a strong course project is expected to satisfy, grouped into four phases from framing the problem through to engineering rigor, each paired with the anti-patterns that most often weaken it.
Project Assessment Criteria
The five dimensions a project is judged on: domain understanding, depth of solution-space exploration, novelty and reuse, methodological rigor, and how clearly the work is presented and defended.
Working Resources
Worked examples, course assistants, and the textbooks the courses are built on.
Past Projects
Completed student course projects with their repositories and write-ups. The most reliable way to calibrate scope and quality before committing to your own proposal.
Course Bots
Course-specific AI assistants grounded in the syllabus and materials. Use them to close knowledge gaps the week they appear, and to challenge your own reasoning rather than to generate work.
Online Textbooks
The open textbooks behind the course series: Language AI, Vision AI, Scalable AI, Temporal AI, and Embodied AI. Readable online, with companion Android builds.