Hacker News Reader

Comment by shockeychap | Hacker News Reader

Throwing up materials that randomly fall into the shape of a crude shelter can certainly happen. But without a LOT of evolutionary mechanisms already in place, that crude shelter will just fall apart, and it certainly won't turn itself into something more sophisticated.

What is necessary for life to arise from a random stew of disorganized chemicals is far more than just a good arrangement of a few molecules that exhibit some characteristics of life. Since all life dies as part of the never-ending march toward entropy, you also need to be capable of some form of reproduction and able to carry out reproduction at a faster rate than death. What is the minimum complexity involved in such a thing? I'd argue that it's probably more complicated than any house in existence.

View on HN

not really, any structure that can absorb energy and replicate could initiate life.

nucleotides spontaneously form, then start to bind to each other in chains. Those chains spontaneously fall apart and reform.

amino acids start to bind to the nucleotide chains and cause them to be more stable. Nucleotide chains bound to the amino acids have a longer life expectancy than those not bound to amino acids. The amino acids start to bind to each other (they naturally do). The RNA binding makes them slightly more energy efficient essentially catalyzing the reaction.

over millions or billions of years, random chaining of nucleotides leads to random chaining of proteins that are slightly more energy efficient. Those forms "outcompete" alternate forms.

Once a particular type of amino acid chain exists, it can influence other amino acid chains to follow the same structure. For example prions are extremely short but binding to normally folded versions of the protein causes the normal version to change patterns.

> over millions or billions of years

one of the interesting constraints is you don't have billions of years. Life appeared on Earth within 100 million years of the planet cooling down enough to have standing water, possibly even less, depending on your reading of the evidence that gets more disputed the farther back you go.

It took 2 billion years for prokaryotes to turn into eukaryotes, and it took another billion for eukaryotes to go multi-cellular, but somehow the initial development of a self-replicating cell with a code took less than a tenth of that.

Reply on news.ycombinator.com