Problem 2. An infinite increasing sequence a1 < a2 < a3 < · · · of positive integers is called central if for every positive integer n, the arithmetic mean of the first an terms of the sequence is equal to an. Show that there exists an infinite sequence b1, b2, b3, . . . of positive integers such that for every central sequence a1, a2, a3, . . ., there are infinitely many positive integers n with an = bn.
SOLUTION 1
Sequence is central if:
Let
.
Becuase
is strictly increasing,
is nonnegative and increaseing.
By some algebraic manipulations we get:
If for some
by the sum above we get
We want to prove that this happens i.o.
If that's true there are infinitey many
such that
and by taking
we are done.
If this doesn't happen i.o.
is eventually strictly increasing. Then sequance
is decreasing.
It can't become negative because of term
in sum. Because this sequance is bounded and monotone it is eventually constant.
Then eventually
for some
and by plugging this in original sum we get
and again
works and we are done.
SOLUTION 2
We claim that

just works perfectly.
First notice having consecutive terms on

would be lovely as if

and

then

which belongs on the sequence so if this happend infinitely many times we are done, else...
Suppose now for all

we have that

then we have to explore how the growth of the sequence itself fights with the rest of components due to the increasingness we are given.
To do this just notice that if at any point for a term

we had that

then it also would follow that

for all

due to the assumption above but then pick some

due to increasingness such that

too and pick a

with similar propeties such that

(yes the excesive bounding here is not needed I just do it for the memes then

however we also have that

which is clearly a contradiction therefore it must never happen that

and thus

for all positive integers

and however this shows that

must force equality everywhere and thus in this case

for all positive integers

and clearly

would also contain infinitely many of these, thus either way we are done
SOLUTION 3
We claim that
holds infinitely many times for any central sequence. FTSOC suppose not for some sequence
.
Define
where
.
Claim:
for sufficiently large
.
Proof: If this wasn't the case, then
holds infinitely many times, contradiction. 
Claim:
for sufficiently large
.
Proof: Fix large
. Note that
Since
, equality holds in the case where
. We can now rewrite this as
Now let
be equal to the nonzero values of
. Then by the above
finitely many times, say
times. Define
. If
occurs at least
times, then we get that for large
,
which is a contradiction since
. This means that there must be finitely many
, say there are
. Then we must have that
for large
, so
. Finally, we have that
as desired.