Current definitions of life find it difficult to normally include viruses and prions without adding sub-definitions, exceptions, etc. Viruses and prions are the simplest known organisms and one idea is to study the functions of the most simple organisms and let them dictate what is life and build a new model and accommodate it in the larger scale. As we know, viruses as living beings exist only in a changing state.
But all life exists only in a changing state and living material can be further reduced and divided till the point we have a single chemical reaction. So , life is a sum of countless chemical reactions. Obviously in the past they were much fewer so we must assume that a source of energy (sun) caused a burst of reactions in terms of number and complexity. But giving energy to increase the number of reactions just leads to mindless chaotic and random reactions. It also suggests that each organism is a system of random chemical reactions, or else a chemical mindless automaton.
This seems pretty naïve as a conception because we know that reactions follow very precise patterns and in fact, can be viewed as being directed by other complex processes and pattern driving structures. But..a) If we consider the whole living system as a unique individual entity, it seems not to have any specific pattern and 
remember who is the reference frame! YOU! Or else a sum of chemical reactions, inside the system which it judges. The cause observed by the result.
After all, what would happen in a growing number of random chemical reactions after billions of years? A) Eventually some sticky reactions would lead to adhesion of molecules that would attract others as well, converting the procedure from diffuse to multifocal, allowing forms to be created, the reactions with repeatability that occur in a somewhat cyclical manner would survive in the long term, because they will not lead to a dead end and c) the reactions that will survive after billions of years will do it because these specific reactions pose surviving capacities over other. From our point of view (perspective) B is perceived as reproduction and C as evolution. What I try to say is that even we were indeed some automaton chemical reactions, even the fact that these reactions continue to happen makes them successful to our eyes regardless of how this happened. These reactions survived and there was a history behind this.
Human position in the system can explain everything. Both life and fire are chemical reactions but fire is very simple with no functional resemblance with us to perceived as life.
Proposed experimental testing:
Due to the fact that even the most simple organisms are very complicating and their precise interreactions with the environment are difficult to be estimated, it is difficult to create an experiment to test IF living beings actually behave as chemical automatons. Here are some ideas though:
A If a living organism is a sum of chemical reactions, then the components of food intake are the first substrates and the excreted products are the last elements. By changing the food and also the pace of feeding, one can observe the way the organism performs some functions, for instance if the organism is an automaton, in certain feeding conditions one can observe extreme outlier values. The latter won’t be observed if the organism is self-regulating (self sustained).
Testing if feeding identical organisms (clones) with the same food in an identical manner and under identical conditions would produce exactly the same amount of waste products plus the error factor ε, or noise, produced by various unpredictable factors. Only if the organism is a system of random chemical reactions, it will behave mechanistically and will produce reproducible results.
The factor ε must follow a normal distribution as known by statistics.
C) If we have clones of the same simple organism and we study them into the same conditions and we give the exact food, then if these organisms are just random chemical reactions, their lifespan could be predicted as a result of multiple linear regression. The dependent variable y (or else the lifespan) would be: y=a+a1x1+a2x2+…….aνxν+aωxω+ε where ε is the error variable and x1,x2…xν the various explanatory variables and a,a1,a2…av the effects or regressor coefficients and aωxω measures the feeding speed effect.
If these clones share everything in common(e.g environmental factors, temperature etc) except the pace with which they are fed and if we secure that actually these organisms absorb exactly the same nutrients, but differ only in the pace they absorb them, then all the parameters of the linear regression will be the same for all clones except the speed factor, or else lifespan=y=aωxω+B+ε (where B=a+a1x1+a2x2+….avxv and it is the same for all organisms), or else we have a simple linear regression. Thus, if we avoid extremes in feeding pace and we assume no collinearities caused by it, then at a certain pace range we would expect lifespan to be linearly correlated with the feeding pace. (ATTENTION: The regressors x do not represent the reactions, but rather represent the effects of some “x” factors. Once again, if the organism is a system of random chemical reactions, it will behave mechanistically and will produce reproducible results. I agree that it is difficult to completely isolate the system from all possible disturbing factors, but if their influence is chaotic and random for all experimental individuals, i think that their influence as a total can be satisfactorily represented by ε , or else the error term or noise in the formula of the final linear regression.
D) One can also test the way the living forms and their functions are decaying when they move to more hostile conditions on earth, such as extreme temperatures, deep ocean etc. Do they decay as if they where random chemical reactions or in an other way, e.g. self-sustaining organsms?
