Which agentic coding tool should social scientists use in 2026?

Video Description

Summary ⤵️ I'm an economics PhD who has been working with agentic coding tools daily for over a year. In this video, I provide a full guide of the different options available for 2026 to help you choose the one which is right for you. Aniket's Socials 🕊️ Twitter/X: https://x.com/aniketpanjwani 💼 EconNow: https://econ.now 🤙 AI Consulting: https://contentquant.io ✉️ LinkedIn: https://linkedin.com/in/aniket-a-panjwani Why watch? If this is your first view—hi, I’m Aniket! I'm an economist (PhD from Northwestern) who worked for six years as an ML Engineer at Zelle, building financial AI models, automations and data pipelines saving consumers from tens of millions of dollars of fraud. Now, I work as an independent AI consultant, and I make videos about whatever interests me on the intersection of econ, tech, and AI. Like, subscribe, and leave me a comment if you have a specific request! Thanks. 0:00 - The most popular tools may be the worst for your research 0:07 - My background: Economics PhD in industry 0:42 - The only three tools to consider 1:45 - All three start at $20/month 1:59 - Why Cursor isn't the best choice for social scientists 2:40 - What Cursor IS good for (frontend touchups) 3:01 - The real comparison: Claude Code vs Codex 3:15 - Bonus: You get Claude/ChatGPT subscriptions included 3:37 - The $200/month tier: ChatGPT Pro wins for social science 4:50 - Plan mode: Claude Code's key advantage 6:59 - Why plan mode matters for your workflow 7:50 - Documentation, context, and the skill gap 8:21 - Using these tools for academic writing 9:06 - Tab completion vs first-draft generation 10:54 - Why you might choose Codex (GPT 5.2 Extra High) 11:01 - Why you might choose Claude Code (skills & extensibility) 12:25 - Final recommendations 13:01 - My current setup