In the 1940s, as digital communications ramped up, Claude Shannon codified basic concepts of how to send and receive information along digital channels. His abstract model of communication goes like this: the message source creates a message in an encoding language, transmits the message, and then on the other end it is decoded by the message receiver.
If there is a message with any coherence to it, there must be a message source.
DNA is an extraordinarily coherent message. It’s encoded in nucleotide bases, decoded by first RNA and then transcribed into proteins, which then fold in 3-D shapes that determine their functionality based on the chemistry of the protein sequence.
A classic argument is the infinite monkey theorem: this is the idea that if you gave a monkey a typewriter, given infinite time, he might at least once by happenstance write Shakespeare’s MacBeth (or War and Peace, or pick your favorite work of literature). Of course, one major problem is that no one argues that there was infinite time–most secular scientists believe the earth is a mere 4 billion years old. The probability of any combination of events occurring in sequence with one another requires multiplying the probability of each individual event–say, a correct first letter out of 26, times the probability of a correct second letter out of 26, times the probability of a correct third letter out of 26, and so on. The probability thus shrinks exponentially. Hamlet is around 130,000 letters. The probability of a monkey typing all 130,000 letters in the correct sequence is thus 1 part in 3.4 × 10^183,946.
To put that in context, there are only 7 x 10^27 atoms in the entire universe, and 10^80 protons (since atoms are made of both protons and neutrons bound together in the nucleus). Even if every one of those protons was a monkey who had been typing at random continuously from the estimated time of the Big Bang (usually supposed to be some 13.7 billion years ago), it would still be statistically impossible for one of those protonic monkeys to accidentally produce Shakespeare’s Hamlet. In order to have even a one in a trillion probability, we’d have to introduce the multiverse once again–there would need to be 10^360,641 universes, each filled with protonic monkeys typing at random for 13.7 billion years (https://en.wikipedia.org/wiki/Infinite_monkey_theorem#cite_note-10).
So for all intents and purposes, the supposition of the Information Theorem is absolutely correct. If there is a coherent message, there must have been a mind to generate it. Information doesn’t come from nothing. It cannot.