In a paper entitled, "On the Very Possibility of Intelligent Design," published in the book, The Creation Hypothesis, back in 1992, William Dembski offered the following thought experiment:
Imagine we find a pulsar (a spinning neutron star), 3 billion light years away, and we notice that it is sending us signals in Morse Code. Further, we realize that these signals translate into proper English sentences. Further, when we send questions back to it in Morse Code, we don't have to wait 6 billion years for an answer. We get one in 10 minutes. [Demski takes the thought experiment further, but I want to stop here.]
Clearly, the signals are intelligently designed. We do not know how they are designed. We do not know how they are sent to us so speedily. We do not know if the intelligence is supernatural or natural. Yet, we know that the signals are intelligently designed. Even without a known mechanism. Even with no way of knowing if the intelligence is natural or supernatural.
So here at least is one case where it is reasonable to conclude an intelligent cause, even though we don't know the mechanism, and whether the cause is natural or supernatural. And it seems that at least part of the reason we know that the cause is intelligent is the degree of improbability involved. It would just be too improbable to continuously receive complex signals that are encoded answers to specific questions.
I readily grant that the degree of evidence of intelligence in the pulsar example far exceeds the degree of evidence we have for intelligent causes of the universe or of biological systems. However, it seems to be the same type of evidence. Which means that determining whether or not intelligent causes are behind the origin of the universe or living organisms is a case of determining the probability of either happening without intelligence.
So even though we may not know the mechanisms involved, or even the nature of the designer(s), we could — if the improbability of other explanations were great enough — know that ID is true. And even though the degree of improbability may not be known, suspecting or even believing that it is an intelligent cause may still be reasonable.