Artificial intelligence is rapidly changing warfare, outpacing the legal and ethical systems meant to oversee it. Global militaries are using AI-driven decision support systems to pinpoint targets, evaluate proportionality, and guide weapons. The divide between technological capabilities and the ability of international law to regulate them is growing daily.
Yuval Shany, a law professor at Hebrew University of Jerusalem and a research fellow at the Oxford Ethics in AI Institute, has served on the UN Human Rights Committee. There, he initially faced the legal and ethical issues presented by autonomous weapon systems, and his research centers around international humanitarian law, human rights, and emerging military technologies.
Yuval joins Matt Merrill for an in-depth discussion in this episode. They explore subjects such as the proximity to fully autonomous lethal weapons, the responsibility gap created by AI-mediated warfare, and insights software engineers can garner from these challenges when developing significant AI systems.
Matt Merrill is a software engineering leader with over 20 years of expertise in building and scaling software teams within enterprise and product-focused settings. His specialization includes backend development, cloud architecture, and distributed systems design. He currently designs and delivers software products as well as leads a team of engineers at DEPT® Agency. More information about his work can be found at code.theothermattm.com.
Please click here to view the transcript of this episode.
Sponsors
turbopuffer is how companies like Anthropic, Cursor, Notion, Atlassian, and Ramp launch their most ambitious search features. turbopuffer is a serverless vector and full-text search engine built on object storage, up to 95% cheaper than traditional search databases and just as fast. With turbopuffer, index and search 50 million documents at 10 millisecond p90 query latency for under $100 a month. Visit turbopuffer.com/sed to enjoy your first month free.
Today’s episode of Software Engineering Daily is sponsored by Unblocked.
Your coding agents have access to your codebase, but access doesn’t equate to context. Agents can’t understand MCPs, your architectural choices, your team’s patterns, or the rationale behind the API’s structure. This results in agents looking in the wrong areas and producing poor output, requiring you to intervene repeatedly.
Unblocked supplies the missing context layer for your agents. It condenses your PRs, docs, Slack, and tickets into organizational context that agents can comprehend, leading to superior planning, better code, fewer token uses, and fewer correction loops.
If you employ Claude Code, Cursor, or any agentic workflow, Unblocked is worth considering. Start with a free three-week trial at getunblocked.com/sedaily.
Every AI team eventually hits the same obstacle. The models are solid and the infrastructure is stable, but the data is hours old because the pipeline is batch instead of streaming, and there’s been no time to resolve it. This isn’t a modeling issue but a pipeline problem.
Estuary provides CDC, batch, and streaming in a single platform with over 200 connectors, operational in hours, not weeks. Your AI’s quality hinges on your pipeline. Visit estuary.dev.