The MU Puzzle

“Gödel, Escher, Bach – An Eternal Golden Braid”

By Douglas Hofstadter
 Kurt Gödel
Gödel is one of the most famous
logicians of all time. He is best
known for his "Incompleteness
Theorem," which proved that in any
axiom system, there are statements
that can be neither proved nor
disproved. Much later in his life, he
showed that the axiom of choice
and continuum hypothesis are
consistent with the axioms of set
theory, which helped lead the way
for Cohen’s proof of the
independence of these propositions.
 Maurits Cornelis Escher

M.C. Escher is a very well known artist.

His lithographs are extremely popular
and thought provoking. These
mathematically inspired woodcuts,
lithographs and mezzotints, feature
impossible constructions, explorations
of infinity, architecture, and
tessellations.

Some of the most famous of this

lithographs are Drawing Hands,
Relativity, Waterfall, Metamorphosis,
Ascending and Descending.
 Escher’s Artworks

Ascending and Descending Waterfall

 Escher’s Artworks

Hand with Relativity

Reflecting Sphere
 Johann Sebastian Bach
Johann Sebastian Bach was a
German composer, organist, violist,
and violinist whose ecclesiastical
and secular works for choir,
orchestra, and solo instruments
drew together the strands of the
Baroque period and brought it to
its ultimate maturity.
Bach’s most important contribution
is his work on Baroque fugue. He is
now regarded as the supreme
composer of the Baroque, and as
one of the greatest of all time.
 The MU Puzzle

"Can you produce MU?"

We start with String MI. We have to generate string MU. At any

stage one or more rules rules can be applied to obtain alternate
strings and the choice of the rule to be applied depend upon the
judgement of the solver. But the string can be changed only within
the rules. This is called “The Requirement of Formality”.
 Rules to be Followed

• A U can be added at the end of any string whose last letter is I.

For Example : MI to MIU I
xI xIU
• String after a M can be doubled. For Example : MIU to MIUIU
Mx Mxx
• A III appearing anywhere in the string can be replaced by U.
For example : MIUIII to MIUU
xIIIY xUy
• A UU appearing anywhere inside a string can be dropped.
For Example: MIUU to MI
xUUy xy
 Decision Procedures
Imagine a genie (or a super computer) who has all the time in the world, and who
enjoys using it to produce theorems of the MIU-system, in a rather methodical
way. Here, for instance, is a possible way the genie might go about it

Step 1: Apply every applicable rule to the axiom MI. This yields two new theorems
MIU, MII.
Step 2: Apply every applicable rule to the theorems produced in step 1. This yields
three new theorems: MIIU, MIUIU, MIIII.
Step 3: Apply every applicable rule to the theorems produced in step 2. This yields
five new theorems: MIIIIU, MIIUIIU, MIUIUIUIU, MIIIIIIII, MUI.
And so on ….

This method produces every single theorem sooner or later, because the rules
are applied in every conceivable order. All of the lengthening-shortening
alternations which we mentioned above eventually get carried out. However, it
is not clear how long to wait for a given string
A systematically constructed "tree" of all the theorems of the MIU-system. The N th
level down contains those theorems whose derivations contain exactly N steps.
The encircled numbers tell which rule was employed. Is MU anywhere in this tree?
 Solution

The answer to the MU Puzzle is No. It is impossible to change the string MI

into MU by repeatedly applying the given rules.
Explanation :
The number of I contained in the strings produced in this case is not
divisible by 3. This is because only 2nd and 3rd rule changes the number of I.
The 2nd one doubles it while the 3rd one reduces it by 3.

• In the beginning, the number of Is is 1 which is not divisible by 3.

• Doubling a number that is not divisible by 3 does not make it divisible by 3.
• Subtracting 3 from a number that is not divisible by 3 does not make it
divisible by 3 either.

Thus, the goal of MU with zero I cannot be achieved because 0 is divisible by 3.

 AI and the MU Puzzle

We can read M-I-U as

M = machine
I = intelligence
U = understanding

Can we produce MU? (Starting with MI) -- This question essentially explores
the difference between Machine Intelligence and Understanding.
In other words, the MU puzzle is really a metaphor for some basic questions
about artificial (that is, machine-based) intelligence. Minds can understand
things. Minds also exhibit the properties of intelligence and can derive
understanding from intelligence. Computers, on the other hand, do not
currently understand things, but can and do behave logically, and therefore
can be imbued with a certain degree of intelligence.
 Continued ..

When enough intelligence is applied, the computer may eventually exhibit output
that is practically indistinguishable from otherwise mind-like (that is, human)
understanding. But would such a display be actual or real understanding?

By using the decision procedures like the one described earlier might take ever to
answer such which might otherwise be trivial for a human mind