Dancing Robot Goes Rogue

This is a headline out of a Hollywood thriller: Dancing robot busts a move, and dishes, after going rogue at California hot pot restaurant. It's an actual story about a robot that gets too close to a table and sends dishes and cutlery flying. In the news report, there is a ticker that says "employees restrain dancing robot in restaurwant.

The robot didn't go crazy or try to hurt anyone. It's really a dancing robot. In reports, it does say that the robot performs a expected, but the humans set it a little too close to the table at the request of the diners. In the video you can also see employee struggling to turn the robot off, which is disconcerting as mechanical systems can be far stronger than flesh and bones. Having kill switches or known ways to disable these systems is important.

Most of us won't likely deal with databases and physical robots anytime soon, but we certainly are being asked to evaluate, use, and perhaps trust automated agents and other AI technologies that may operate on our code and our live systems. More and more organizations are looking at AI to not only examine data and provide code, recommendations , or analysis, but with agentic systems, they may take actions.

What happens when a "digital robot" goes haywire in a software system? What if a monitoring system starts killing off sessions because it "thinks" (really predicts) that they are causing slowness. What happens if a DevOps robot thinks your code is ready and starts deploying changes? Or maybe worse, what if it detects issues and starts issuing undo commands for the last few changes made to the system? I know we've been concerned about these sorts of actions, AI-based or not, and tried to be careful what capabilities we build into Redgate Monitor. Adding too much autonomy could easily cause more issues than it solves.

There is tremendous opportunity with robots and agents, but we need guardrails, strong controls, circuit breakers, and easily accessed kill switches. While many of us would love AI to do our laundry, what happens when one of those robots goes wild? Stained clothes might be the least of our problems.

Steve Jones - SSC Editor

Join the debate, and respond to the editorial on the forums

Get AI-Ready With Erik: Retrieval Augmented Genera...

Discover why many enterprise AI projects fail and ...

Early in Covid, even before the mass awareness set...

Lately, I’ve been feeling like we’re not paying as much attention to transaction logs as we should. In fact, I’ve been saying there needs to be a flashing neon...

When you take a backup, the allocated extents are read from disk and written to wherever you asked, this could be local disk, a network location, some place in the cloud across the world. Seems pretty straight forward, taking away all the compression and encryption stuff for now. However, the database probably has things changing it while the backup is going

Learn what the SQL Server Resource database is, why it exists, how it affects patching/ upgrades, and what to do if it’s missing/corrupted.… The post Everything you should know about...

In a previous post, I covered what SQL MCP Server is, why it matters, and why getting the security model right is non-negotiable. If you haven't read that one,...

On March 10, Microsoft dropped a security patch for SQL Server 2016. It fixed a publicly disclosed, actively scrutinized elevation-of-privilege vulnerability — one that lets any authenticated user escalate...

I’m going to open with a perhaps controversial statement: “when you buy 4 vCores on the Azure SQL Managed Instance platform, what you’re actually buying is 2 physical cores presented as 4 hyperthreaded cores to SQL Server”.

So, I wrote a blog about getting Mental Rest on Si...

(Warning: This blog post is non-technical but pure...

I was listening to the “excellence, actually” podcast recently, and one of the co-hosts mentioned an idea that I recognized but had not heard termed in such a way...

Part 3 in a series on evolving SQL Server environments into AI-ready architectures. Once organizations recognize the architecture gap we discussed in a previous post, the natural instinct is...

We’re coming back to New York, which is exciting for me. I love NYC. The Redgate Summit 2026 – New York City comes back on May 5, 2026. You... The...

The third edition of the Mastering DAX video course is available! The third edition of the Mastering DAX video course is now available! The world has changed a lot since the release...

There are countless project managers and finance assistants all over the UK building summary dashboards, all doing the same thing.

Yesterday, I gave a presentation at Microsoft Fabcon in Atlanta titled “Building the Well-Architected Framework for Fabric”. I’ve been a public speaker for nearly 20 years, and I don’t usually get nervous before a talk. Not that I was shaking yesterday, but given this was a new topic for me, and a big crowd

I have been spending a little time with the Microsoft Fabric data agent documentation lately, and one pattern keeps showing up, and it is not just in the official guidance but in community posts from people who have actually tried to deploy these things: the demo runs beautifully.

I previously wrote about how the underlying technology for Fabric mirroring changed with SQL Server 2025. The latest version of mirroring that uses the SQL Server Change Feed is... The...

What Microsoft just revealed about the future of enterprise analytics… Microsoft’s FabCon 2026 conference unveiled a transformative vision for enterprise data and AI integration. Rather than incremental improvements to...

PasteThePlan has a new AI Suggestions tab, and it’s been really fun to watch the query plans come in and the advice go out. Here are some examples: Date...

There are a huge number of best practices around SQL Server performance tuning – I could easily write a whole book on the topic, especially when you consider the...

Learn PostgreSQL User-Defined Types (UDTs) with clear examples of DOMAIN, COMPOSITE, ENUM, and RANGE types, plus practical SQL code for real-world use.… The post How User-Defined Types work in PostgreSQL:...

Introduction: The End of “Just SSH Into the Box?...

Welcome to Part three of our series for building a...

Recently during one of the Oracle to PostgreSQL mi...

Introduction Most people who work with Postgre...

Your PostgreSQL HA cluster promotes a new primary....

For Power BI developers, one very common (and frustrating) issue is when measures spill into future dates on charts especially when working with some time intelligence DAX calculations (e.g. MTD, YTD, etc.), date dimensions that extend beyond current date, and forecast-enabled tables.

Video by: Reid HavensReport-scoped measures (aka r...

Implement a coding agent in 131 lines of Python code, and a search agent in 61 lines

A while back I posted about a couple of side projects that I’ve been working on when I get the chance. One of those was the Burrito Bot…a bot... The...

Someone runs a massive SELECT INTO #temp, tempdb f...

There are a lot of shopping days left before Christmas, and even more before the next version of SQL Server ships, so might as well get my wish list...

BCP. I love it. It's a workhorse — bulk copy data in, bulk copy data out, no drama. I've used it more times than I can count. Which is...

In cryptography, the RSA and ECC algorithms which we use primarily for asymmetric cryptography are susceptible to Shor's Algorithm in quantum cryptography. Cracking those algorithms with classical computing is... The...

Merge conflicts don't have to ruin your day. I use mergiraf for structural merging and a Claude Code skill for everything else. Here's how to set it up.

A month and a half ago, I wrote a blog on using DATE_BUCKET. It is a cool feature thta makes doing some grouping really quite easy. It is here:...

During the Erin Stellato and Makena Barickman sess...

Copilot in SSMS has two parts. Usually people focu...