gps372

This might be already happening. Competitor's market share (without details) is common to share. Scintilla onboarded by Walmart earlier this year might already be giving instore AI capabilities.

True, in-store planogram compliance and floor-plan compliance have got in focus again. It never lost the focus to be honest, just that now there is an expectation to do it while ensuring nothing fall through the cracks.

But there is a much more stringent privacy requirement here since customers walking in those stores haven't really signed up for this analysis. Hence, difficult to just setup the camera and take continuous data for analysis.

Plain and boring computer vision would have still required a team who would identify frames to sample before processing, and then annotate, generalize and classify the data.

Not an expert on this but I don't think same kind of manual effort is required anymore.

Most comments may have ignored this so far, but this is an interesting AI based research.

At the end of the article is this excerpt

>> Results from this research were generated using an AI computer vision model developed by the research team on Bristol’s Isambard-AI, the UK’s most powerful AI supercomputer. Isambard-AI powered the analysis of 172.6 hours of live match footage from all 104 games played between 11 June and 19 July 2026.

So, basically now merchandizers, auditors, compliance monitors, etc. have a tool using which they can quantify whether broadcaster complied with contractual requirements of showing their ads/merch/logos as per contract. And if they missed, then they can quantify the delta - to sue the broadcaster for the balance as well.

Also, it can enable many countries, where certain content is not allowed, then this tool can simply give them data points without having to employ someone, or have someone review the output of the AI tooling.

Hmm, in your example, temperature is fully measurable but there might be a slight challenge in measuring if "If we are at home". Since for this you need to define more variants with measurements from motion sensors, pressure or weight sensors on beds or sofas, if a device has got connected to wifi, etc.

By the time your first round of beta testing is over, you may have quite a handful of such axioms and variants, which have been humanly validated!

Have you done any such experiment in this space?

Haven't gone through full PDF as its very detailed, few things have resonated with me so far.

Basically if a Car A is performing better (be it speed, milage or in general sense) than Car B, then it is not necessarily because its engine. It could be because of better tires, better gearbox, lighter body, better usability of features, etc.

You can implement an AI feature (like AI for BI) in different ways even with the same model - via ReAct-loop, or plan-and-execute, or hybrid. You can make it stateless, stateful, RAG-based, etc. depending upon whether you want to prioritize result accuracy or depth of analysis. You can use LLM to generate either intent (requires lesser reasoning) or the queries itself (requires much more capable model).

Your harness can adapt to the underlying model's native capabilities, or can make up for its absence, e.g. query generation in above example requires your model to have MOE capabilities but intent generation wouldn't.