An important input to technological development is economic productivity. From Nick Bostrom's The Future of Humanity (2005),

One could argue that to get an index of the overall pace of technological development, we should look not at a hand-picked portfolio of hot technologies; but instead at economic growth, which implicitly incorporates all productivity-enhancing technological innovations, weighted by their economic significance.

Furthermore, many proposed AI takeoff scenarios are framed in terms of economic growth and development (See for instance Baumann and Hanson). Economic investment has also been pointed to as a driving factor of the recent AI and compute trend.

If Paul Christiano's slow AI takeoff view is right, then economic growth will speed up long before we have advanced AI technology. That's because pre-AGI technology can potentially automate labor without strictly being smarter than all humans. This would imply that upon the arrival of AGI, the world economy will be much larger than it would be if you naively extrapolated from past economic growth.

For example, suppose you thought that AGI was 20 years away, but that 10 years before AGI, economic growth would be double the current rate, triple 5 years before, quadruple 2 years before etc. Then although the GDP's current instantaneous growth rate implies that GDP should be about twice as much when AGI arrives, it will actually be much higher due to the pre-AGI-induced acceleration. But exactly how much larger will it be?

The relevant year when AGI is deployed is determined by this question. A real GDP estimate for that year from a reliable source will determine the resolution for this question.

Real world GDP was estimated to be about 87 trillion dollars in 2019.

ETA: The question resolves as real GDP in 2020 USD.