10 tips to level up your ai-assisted coding - Aleksander Stensby - NDC Copenhagen 2026

Value of the Information & Strength of the Argument

The talk provides high-value, actionable advice for developers using AI coding assistants. The speaker’s arguments are grounded in practical experience and align with current industry trends, such as the adoption of MCP and skills. He effectively argues for a proactive approach to managing context and rules to achieve compound engineering, where AI improves over time. The emphasis on skills as portable assets is particularly valuable, as it addresses vendor lock-in concerns. The argumentation is coherent and persuasive, though it relies on anecdotal evidence rather than empirical data. The speaker’s credibility is enhanced by his background in machine learning and his hands-on experience with Claude Code.

Scientific Rigor, Source Quality, Title Accuracy

The talk demonstrates scientific rigor in its practical approach, though it lacks formal citations. The speaker references concepts like compound engineering from the company Every and mentions Boris Cherny, the creator of Claude Code, but does not provide specific sources. The title accurately reflects the content, which is a list of ten tips. The talk is well-structured and the advice is consistent with current best practices in AI-assisted development. However, the lack of external references and empirical validation slightly reduces its scientific rigor. The speaker’s practical experience and alignment with industry standards partially compensate for this.

The title accurately reflects the content, which presents ten categories of practical tips for improving AI-assisted coding workflows.

8/10

The talk is based on the speaker's extensive practical experience with AI-assisted coding, and it aligns with current industry practices and emerging standards like MCP and skills. The advice is pragmatic and actionable, though it is primarily anecdotal and not backed by formal research or empirical data.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 02:00Introduction: Speaker shares his background and the shift to AI-assisted coding.
• 05:00Tip 0: Mindset shift - collaborate with AI, not just autopilot.
• 10:00Tip 1: Context is king - guide AI with context, use visual inputs.
• 15:00Tip 2: Manage context window - fresh glass, control compaction.
• 20:00Tip 3: Use rules and memory for compound engineering.
• 25:00Tip 4: Invest in skills - portable, deterministic, and powerful.
• 30:00Tip 5: Always start with a plan - use plan mode and describe steps.
• 35:00Tip 6: Be specific but not too specific - avoid limiting AI's creativity.
• 40:00Philosophical reflection: The bottleneck shifts from coding to describing and evaluating.
• 45:00Q&A and concluding remarks.

• NDC Conferences — Conference organizer and host of the talk.
• NDC Copenhagen — Specific conference where the talk was given.

• Model Context Protocol (MCP) — The talk discusses MCP as an emerging standard for connecting AI assistants to tools, aligning with this official resource.
• Claude Code documentation — The talk references Claude Code features like skills and plan mode, which are documented here.