How AI Growth Could Impact Oil and Energy Markets

If AI is really going to be adopted as fast as people say, is that not very deflationary? If deflationary then how would oil/energy markets be affected?

• Deflation: a decline in prices for goods and services.

• Widespread AI uptake would most likely be deflationary on economies, as big increases in productivity because of AI would mean the output of the economy exceeds the supply of money/credit. In real terms, this means the cost of output would be brought down by AI efficiency, and the size of output would also increase. For example, 1gb of data in 1980 cost $437,500,  and by 2014 it cost 3 cents. 

• Rapid and large deflation is not a good thing, as the lowering cost of goods means higher unemployment (not everyone can still be paid for a company to be profitable), and naturally many jobs would be lost anyways to AI automation (as in many cases it would probably do a better job than a human).

• Query: Would real jobs be sacrificed on a massive scale and replaced by AI, even if it is the logical or economic thing to do, and if so, to what extent will this happen? Will there be government / company mandated policies to prevent this, or even a general human consensus against taking people out of jobs, as this could lead to big unemployment issues?

• There are still lots of people who are anti AI or prefer the human component in work. For example, I have heard of law firms and their clients completely banning AI use even for back office tasks due to a lack of trust in the diligence….

• In the more likely case AI becomes heavily adopted (not necessarily completely), deflation would likely occur. Even as prices come down, spending should also decrease because of unemployment increases, and people are more likely to save when costs of goods are consistently declining.

So, what does this mean for the oil/energy markets?

Oil:  

• I would think as economies become more productive because of AI, you would expect a larger demand for oil products as businesses start to take advantage of greater outputs at lower running costs. In the case of oil, this could be seen in the full automation of transportation, manufacturing, and agriculture. 

• However, there is a case to be made that automation in these industries is already reached or close to being reached, so the extent of this increase in demand for oil because of greater productivity from AI might not be great.

As deflation effects starts to kick in, I would then think prices for oil would go down with a general decrease in costs for goods and less people buying them.

Further Reasons:

• Oil may be used for some power generation, but this is in decline and only used in some economies, with other energy sources being the main drivers in electricity generation (natural gas, nuclear, sustainable alternatives).

• Even in transportation/manufacturing/agriculture, there is a general trend towards electrification so this initial demand for oil in these sectors may not be great anyways (depending on when AI really does take over).

Electricity does not = oil:

• Datacenters & GPuS: Training and running AI systems will increase electricity demand (coming from sources other than oil).

• Manufacturing robots & automation: Increases industrial use of electricity, less so oil.

• Small caveat: Energy demand may not be perfectly linear as ongoing hardware/software optimisation may reduce energy needed per computation. 

Putting it together in the long-term:

• AI = likely deflationary. 

• Cost of goods driven down, lower spending activity & higher unemployment in deflationary environment —> ultimately keeps oil prices down.

• As electrification from other sources is greater than any oil-fired electricity - AI impact on oil consumption is limited. This should continue to be limited.

Potential upside for oil:

• If AI-driven growth is very rapid, overall demand for all forms of energy may still climb, especially if supply (e.g. for natural gas or renewables infrastructure) can’t keep up in the near term.

• Oil will still likely be used in industry, considering the growth of AI will likely happen faster than the world’s shift away from oil use. Markets will still be influenced by big geopolitical factors or OPEC+ decisions - so even an otherwise deflationary environment might see episodic oil price spikes.

• AI growth may not reach all countries in the world, and developing economies will still be reliant on oil in industry. 

Oil prices likely see an uptick and then extended downwards movement (like a rollercoaster), energy prices overall increases

Note: The biggest caveat of them all is that nobody knows the future. Perhaps AI will not develop and be implemented as fast or as widespread as this article considers!