Average Is All You Need

The claim

It has been reported that RawQuery published an editorial arguing that "LLMs have eaten the world" — and that’s okay. The piece, titled "Average Is All You Need," makes a quiet, contrarian case: large language models have democratized the middling stuff. Average writing, average visuals, average analytics — all suddenly cheap, fast, and available to anyone who can type a sentence. Sound bleak? Or liberating? For many, the emotional hook is relief. You no longer need to be a specialist to get something useful out of your data; you just need to know what you're trying to learn.

The demo

Allegedly, RawQuery's product is built to let LLM agents do the grunt work: connect data sources, generate SQL, run queries, and render charts. The blog walks through a concrete example — hooking Stripe transactional data to HubSpot campaign data, syncing both sources, then asking an agent (they name-drop Claude Code and Cursor) to test whether an email campaign raised average basket size or customer count. The code snippets in the post show commands like rq connections create and rq connections sync, followed by schema and table checks. The pitch is simple: describe the join you want in plain English; the agent writes the SQL and returns the chart.

Why it matters

This isn’t just a demo; it’s part of a broader trend where AI handles the average so humans can focus on the hard, messy thinking. That’s the upside. The downside? Data quality, attribution choices, and security still bite. Attribution models were already described in the post as “wankery” — a funny word, but a real problem — and automating them doesn’t erase the ambiguity of causal claims. It has been reported that RawQuery expects these LLM-driven workflows to lower the bar for useful analysis, freeing teams to iterate faster. Whether organizations treat that as progress or as another source of false confidence will be the story to watch next.

Sources: rawquery.dev, Hacker News