Kinematic Self-Replicating Machines

© 2004 Robert A. Freitas Jr. and Ralph C. Merkle. All Rights Reserved.

Robert A. Freitas Jr., Ralph C. Merkle, Kinematic Self-Replicating Machines, Landes Bioscience, Georgetown, TX, 2004.


4.8 Laing Molecular Tapeworms (1974-1978)

Starting in 1974, Laing [557-563] proposed a hybrid cellular-kinematic machine replicator system which would compute and construct by means of a repertoire of simple mechanical actions (e.g., sliding local shifts of contact, local changes of state, and local detections of such state changes). The system would manipulate componentry of the same sort of which it is itself composed, and would in essence comprise a molecular tape that replicates, making more tapes identical to itself, using a universal Turing machine motif (Figure 4.28(A)). Machines of the system can imitate the actions of any Turing machine or of any robot system like themselves [558]. Notes Laing [557]: “We focus on artificial macromolecular machines. In form these macromolecular machines are chains (possibly folded and interconnected; Figure 4.28(B)) of basic molecule-like constituents....a molecular machine, an automaton device all of whose basic constituents and operations are the sort that reasonably may be said to be possible for biological systems at the macromolecular level....both machine and description will observe biological reasonableness.”

The Laing kinematic automaton system is related to both the cellular automaton system and the projected kinematic machine system of von Neumann, as well as to the basic Turing machine concept [557, 559]. In the Laing system, connected strings of finite state automaton componentry interact by making sliding contact and sending signals into each other. In this system, new components are recruited from the surrounding space at the ends of the interacting strings. These kinematic automata strings can be viewed as a mathematically precise analogy to certain biological processes at the macromolecular level and “might better be called molecular machines.”

In one version [557] composed of rigid constituents (Figure 4.28(C)), Laing’s kinematic replicating system consists of a program chain interacting with a tape chain. The tape chain is envisioned as a string of molecules of a sort able to take on two distinct physical states (e.g., 0 and 1), “which might in a physical molecule correspond to the presence or absence of an auxiliary molecule or to one or another of two distinct orientations of a molecule type.” The program chain also can be represented in a string of effector molecules, and there are ten basic instruction types (requiring 10 different types of effector molecules): (1) write zero or “W(0)”; (2) write one or “W(1)”; (3) move left or “L”; (4) move right or “R”; (5) halt or “H”; (6) no operation or “NOP”; a branch instruction consisting of (7) an origination conditional transfer or “CT” molecule and (8) a destination transfer target or “TT” molecule; (9) a “detachment” molecule which when “activated and in temporary association with a molecule, will cause the severing of the structural attachment that molecule has with its predecessors”; and (10) a “synthesize” molecule which upon “being brought into active association with a passive molecule (a parts blank) will cause the passive molecule to change its state.”

In the operation of the molecular machine, a string of program molecules is in sliding contact with a tape molecule string at a single location. If the active program molecule is a W(0), it will affect the contacted tape molecule to place it in state 0; if a W(1), to place the tape molecule in state 1; if an R, to produce a slide to the next tape molecule to the right; and so forth. If an R or an L would result in disengaging the program and tape molecule chains, then a new “0” molecule must be synthesized or otherwise be recruited from the environment and attached to the end of the tape molecule string.

The molecular machine, composed of program chain and tape chain, then replicates as follows [557]: “First, the molecular machine produces a string of molecules which is a description of itself. Referring then to this description, the molecular machine determines one-by-one the type and constituent required, then retreats to the end of the description string and synthesizes the required constituent. By continued repetition of this process the molecular machine will finally end with a sequence of molecules, the first half of which is its description in ‘0’ and ‘1’ molecules, and the latter half of which is an exact copy of the molecular machine constituent primary sequence. The detachment molecule can now be employed to sever the connection between the description molecule and the new molecule machine constituent sequence. We shall also want to be able to communicate an ‘active’ status to a region of offspring molecular machines, else offspring would possess no autonomous self-reproducing properties. During construction, and upon release, the presence in the primary constituent sequence of structure-forming properties will by self-assembly impose the requisite three-dimensional form.”

More recently, simpler versions of self-replicating molecular strands or strings (e.g., “typogenetics”) have been introduced by Hofstadter [418] (as a formal way to describe operations on DNA strings) and further discussed by Kvasnicka and Pospichal [2302, 2303], Lynch [2304], Morris [2305], Snare [2306], and Varetto [1659, 2307]. Hutton [2308] has performed computational simulations of artificial replicators that inhabit a virtual soup and are subject to simulated laws of artificial chemistry; see also our discussion of the JohnnyVon simulations in Section 2.2.2. The interactions of computational objects forming a “Turing gas” in a form of “algorithmic chemistry” have been investigated by Fontana [597-599]. Hatakeyama and Nomoto [2309] have used a “movable finite automaton” model to simulate the synthesis of a copy strand of mRNA from a (stringlike) sequence of DNA. Ikegami and Hashimoto [564] have also described a network model of self-replication involving machines and description tapes.

Finally, futurist William Atkinson [2310] while skeptical of machine phase nanotechnology [208] nonetheless believes that molecular assemblers based on “nano cellular automata” could produce copies of themselves or other structures made of diamond and other materials, possibly using a more organic or nonmechanical process.


Last updated on 1 August 2005