I would like to come in to add a little bit of colour. This is potentially a completely transformational technology and, from a productivity point of view, it might be incredibly good news for the economy, as opposed to a lot of the bad news that we have just been discussing.As your question indicated, there are two aspects to the AI issue. One is about how we think about it, and the other is about what evidence we take on it.How we think about it is exactly as referenced in your question. There is an AI-producing sector that is producing hardware and writing all the software that we are downloading and using, and then there is an AI-using sector, which is made up of companies, airlines, the health service and whoever it might be, that will use all that software in providing the services that we all want.Callum Bridge-Adams was referring to our sectoral analysis, and I will add a bit of colour to that. If you look at the industry breakdown, you can somewhat identify what makes up the AI-producing sector and the AI-using sector. The AI-producing sector is the information and communication services industry. It is writing the software; it is writing all the clever bits of code that the lawyers are using to automate their opinions and all that kind of thing. The AI-using sector is the health service, the retailers and all that kind of thing. That is stage 1, which is how to think about it, and that is exactly in the way that your question indicated.Stage 2 is to look at the data, which we do in table 2.2 of our report. I will say a couple of things based on that data, if I may. One is that we have in the UK—I include Scotland, obviously—a pretty strong AI-producing sector. We have a very strong university sector; I do not need to tell you that the University of Glasgow, the University of Edinburgh and other excellent universities are here. Our university sector is strong on computer science, maths and so forth, so the AI-producing sector looks rather large in the UK relative to other countries. It also has pretty high productivity growth—we have numbers around 7 per cent and 8 per cent for the AI-producing sector broadly defined, which are very large numbers. The issue is that, in order for that to have a material effect on overall productivity, everybody else has to use the AI. That is the dilemma, I think, for Europe as a whole.If you look at the numbers showing how much the software is being used in the AI-using sector in America, you see that AI use is booming in the retail sector, the health system and so forth. If you look at the numbers for Europe, you see that they seem to be rather lower. There are lots of measurement difficulties around all of that, which the Office for National Statistics is looking at. I hope that we will be able to say more about the use of it as time goes by and we get better data and better indications from that.I am sorry if that was a rather long answer to the question, but it is a crucial technology. That is how we are trying to think about it and to monitor what its knock-on effects might be.