Eventually we will hit the point at which oil prices do go up forever -- if they are measured as the inverse of EROEI. In other words: we'll reach the point at which every barrel of (conventional geological) oil that we produce takes more energy to extract and process than the last. At that point, oil is a dwindling resource.

This may or may not mean that the price of oil will go up forever in nominal dollars. It may actually have the reverse effect at times, since it could trigger financial crashes that result in deflation. It may also lead to replacement of oil with other resources (gas, electric, etc.) or shifts away from oil-based transport (electric trains, walking, biking) -- and those could lead to decreases in oil price as well if they occur to enough of an extent.

The error in predicting eternal oil inflation is threefold:

(1) Assuming that financial cost always reflects objective physical cost -- that there is a 1:1 relationship between thermodynamics and price. Reality: there's only a soft relationship subject to the next two factors.

(2) Assuming infinite demand inelasticity. Reality: increasing oil price fuels substitution and demand destruction.

(3) Forgetting that money (especially fiat money) is itself variable in value. Reality: deflation can cause prices to nominally fall when nothing has actually changed, and inflation can cause the inverse.