I appreciate that he's waiting with bated breath and I appreciate that he has an interest in all of this. To the point, though, this is not about wasted time. This is about answering the questions that we had kind of put toward witnesses and weren't able to continue on with in the study. It's about trying to make sure that we outline for everyone here why it's so important that we go ahead with the electricity study.
When we're done with that study, let's go to Mr. Angus's study, a study that I've said I agree to. I am good to go ahead with that study. I will probably be proposing some amendments, as I said, to make sure we have community voices included. But we had a study before us. We had a panel before us ready to answer our questions. We had to dismiss them before the normal time. We had a project planned for this study. We haven't been able to continue with it and we won't be able to continue with it if the amendment proposed by the Conservatives goes ahead.
I'm just hoping that by outlining some of these examples I'm helping to convince the other members of this committee why it's so terribly important that we do look at this study on the electrical grid and that we don't just toss it aside.
Before I talk a little bit more about the low-carbon economy fund and other kinds of funding supports that we have for electricity that our government has put in place, I would ask whether this is something that the other members of this committee would agree to: Let's not go ahead with this amendment. Let's drop this amendment. That's a possibility. Let's complete the clean electricity study we had in front of us. Then let's move on to Mr. Angus's study.
The interesting part about this is that, as far as I can tell, although I don't think I've heard from absolutely everyone yet, not only would Mr. Angus's motion pass; it would pass unanimously, I believe. Why not take that opportunity? It would actually be a rare and beautiful moment for this committee to say, hey, we're all agreed on this. We all want to study it together. We might be coming at it from different perspectives as to what we want to bring through with Mr. Angus's study, but let's do it. Let's do that study.
Why supersede the electricity study? Why do away with the work that we've started? Ultimately, by the way, every time we do something where we start a study and then we stop it, and we stop it for a whole period of time, and then we get back to it—it could be months later or almost a year later—it just ends up being so outdated. It's outdated from what we've heard from the witnesses we've had already. That becomes dated. You have to call them again. It means that things overtake you, to some extent.
I've been talking about the importance of planning ahead and looking ahead, and about all the steps of what we need to be looking at, be it from the employment perspective, be it from the investment perspective and be it about the different regional needs. We either get to complete the study now and make it a timely study, making the evidence that we've already heard timely, or we lose that. It becomes an outdated study that we complete a year from now. That's what I'm really imploring the other members of this committee to think about.
Why not just go ahead, have all committee members show that kind of co-operation and goodwill and agree to Mr. Angus's study? I believe that motion will pass once we get to it. The challenge is that the Conservative amendment that was brought would supersede the electricity study. That's the part that doesn't quite make sense to me.
That's what I'm hoping for.
If it wasn't about all of the statements I was able to bring forward until now about budget 2023.... If they don't convince people and get them excited about what we could be studying in the growth opportunities, the jobs and the affordability pieces.... If they don't do the trick, I have a few other things I can bring to mind that might help people think about this a bit more, and about what they would like to see and do.
Before Mr. Angus's intervention, I was talking about the low-carbon economy fund, which provides some very interesting and different funding projects.
I was talking about my alma mater, the University of Toronto. This was interesting to me, but it makes sense when you think about the size, not only of all its buildings, but also of the number of students and faculty on campus. There are so many people. It's really like a small town. The university is built a geothermal system right in the centre of its campus. It's a district energy system. It can actually help support the City of Toronto system a bit if it has excess energy along the way.
The low-carbon economy fund helped it do that. That is a great support to reduce emissions.
I'm going to have to look into it again, but I think the University of Toronto has been rated one of the top universities from a clean energy or environmental perspective. It's also building tall timber, but that's for a whole other study on another day.
Anyway, that was just something I wanted to flag about the low-carbon economy fund in the targeted programming. That's the top of the pyramid when we're looking at clean electricity. It's all those building blocks that fit together.
I hope this goes to show that, in fact, Canada has been doing a lot, and our federal government has been doing a lot, to put in place all of these building blocks to make sure that we have what we need to get that strong, affordable, reliable and clean electrical grid.
That is some of the stuff you have seen.
Now, going to the other piece, because I've been talking about economic growth, one of the parts I would be really interested in being able to ask more questions about, when we talk about electricity and our electricity needs, is the impact of artificial intelligence on our energy needs as a country. When we look at this budget, budget 2024, there are investments in artificial intelligence. It's a growing sector. It's certainly a sector that we are trying to grow in our country. It provides a lot of opportunities, again, for different kinds of good-paying jobs. It's a different part of the economy that we're growing.
I also talked a bit about how the auto sector investments in battery technology, manufacturing and all of that bring with them needs for more energy, but what I didn't really focus on, and what I would really like a chance to ask witnesses about, is the impact of that. What do we need to be thinking about around AI and our electrical needs? Again, it's probably going to have regional implications.
When I looked at that, I was reading some articles and just getting some information about it, and there was some work on what is even a need to have Energy Star ratings for AI models.
I didn't even realize that training GPT-3, for example, is estimated to use just under 1,300 megawatt hours of electricity. That's about as much power as is consumed annually by 130 homes.
This is kind of a fun comparison. Streaming an hour of Netflix requires around 0.8 kilowatt hours of electricity. That means you'd have to watch 1,000,625 hours of Netflix to consume the same amount of power that it takes to train GPT-3. That's from an article in The Verge that came out earlier this year.
That's a lot of energy that we're going to need when we're looking at AI. There are huge possibilities, good-paying jobs and an area of growth for our country, but that's going to have a different kind of regional impact.
Sometimes we ask questions and we think about it in terms of some things that we've traditionally thought about, like certain manufacturing sectors and what those electricity impacts are. We can ask about what it means as we talk about changing the way we move our vehicles, the way we heat and cool our homes, but the other question is, as we build in new technology and industries, what energy storage will we need?
As an example, there was a paper by Dr. Sasha Luccioni, who is a leading AI researcher and climate lead at Hugging Face, based in Montreal. She's the one who had advocated to introduce Energy Star ratings for AI models.
She was talking about a test of 88 models generating text versus image generation energy uses. If you were going to use some examples of what that looked like for images images, it was based on 1,000 requested with text versus images. On one little test, the amount of power was equivalent to running a washing machine for about 2.9 loads of laundry. I know a whole lot about laundry. I do a lot of laundry in my home. It's an unfortunate thing, but there you go. That's a whole lot of laundry that you get for just that one image.
On projected growth in energy use from AI, Alex de Vries, a Ph.D. candidate at the University of Amsterdam, uses Nvidia GPUs to estimate AI's global energy usage. That currently represents about 95% of the AI hardware market. It provides specs and sales projections.
The calculation—