Unlike many mastodons, I have never read anything Elon Musk has
produced. I had a friend who was a fan of Joe Rogan podcasts, when Joe
Rogan was less objectionably pigeon-holed and heard a little bit of
that.
Elon musk does own the bird place, and I am reading and appreciate their
message passing stream deep learning research, though this has a back
seat to my own personal-scale hopfield nets at the moment. Elon came to
own the bird research.
That's my excuse for not knowing if Lonny is more evil than I appreciate
here. I hope he's not. But I think we can agree even his attempts to
save face have lacked grace, mildly. As I am not a primary source here,
I leave off criticising, but the good news for Elon Musk personally is
that I've got solutions he can execute that are even on brand for him.
Elon Musk is associated with the Paypal move of giving out $10 coupons
to get people to sign up for accounts until Paypal was popular enough
that ebay would acquire it (ugh, financial maneuvering).
Right now we are listening to the swan song of large-data-purchase
language model chatbots by Microsoft and copycats, based on the former
public research of now-Microsoft OpenAI, owned by Peter Thiele, who is a
slightly different person to Elon Musk.
Despite my extremely positive community experiences with seasonal fruit
picking, my final thesis is that low wage employment is in primacy a
tool to keep massive numbers of people exhaustingly busy not doing
anything threatening to the encumbent capital in our capitalist society.
This is a hypothesis itself- but read on, these topics coincide and Elon
Musk could be the one to change the world.
Given that there's money in spades (ignoring that less than 1% of people
has the money) and not enough work now and in general, there is a new
and previously impossible strategy which Elon Musk has previously
endorsed but not realised yet.
One of Elon's ideas is supporting humans fusing with technology, rather
than just buying a paid subscription to it as a web service.
So we can do two important things that are out of vogue:
(1) People have software, including code, on personal devices they
carry, that is run wholly on devices they carry, with data they carry.
(2) All people need a small fundamental income, such that they could
choose to act of their own volition rather than do consuming corporate
busy-work to pay off rent seekers and eat food. There's nothing
normative here: Simply this new choice will be open, which it is not
currently.
Well, there's so much good news in technology with rampant single board
computers that aren't locked into being dumbed-down smartphones and one
day affordable repairable laptops that (1) has already happened for many
of us, and if (2) is fulfilled, is more open to more of us.
What Lonny can do is (2). Since governments are the glove part of the
hand-in-glove of corporate greed, it seems unlikely governments will
support freedom from rentseeking and hand-to-landord's-mouth
wageslavery. Instead Elon Musk can pay people ("hire them" if you will)
taking absolutely no obligations from the hire-ee (Elon Musk won't claim
to own their thoughts, actions and future actions, despite paying them).
But what's the return to own on this? Which brings us back to (1).
Simply parking a chatbot fuelled by megacorporate data purchasing
budgets and cheap electricity in the front of an Answer Website Business
or other Web Searcher has been a giant business success. The nature of
these desirable results is that, like a skilled StackOverflow sewer
expert, this experienced spelunker can produce somewhat desirable expert
results in fields they are not an expert in. Optimised non-linear access
to its input knowledge is broadly what deep learning is and does. These
expensive, secret-data-set trained chatbots accelerate users by this
power.
The way this power has been used is a techno-bio-phase-space swing and a
miss, ironically by its prohibitively expensive nature, which Microsoft
has announced is a dead end. Microsoft (and alike) have purchased and
groomed the world's surveillance capitalism such that their chatbot can
be spelunked to produce sometimes operable expert answers (albeit not
great ones (think security coding examples) - but from a capitalist
viewpoint they might be called good). By good, I mean people's
productivity is at a much accelerated velocity than it otherwise would
be
except that everyone is signing up to one (okay, a few equivalent)
helper chatbots. We have a population of high-velocity, accelerated
users of the small number of different deep learning trained chatbots. I
am about to argue that since everyone is being accelerated in the same,
small number of dimensions, the world-changer is that increasing (2)
feeds back to (1).
Instead of corporate employees being accelerated in a small number of
subscription helper chatbot dimensions, accelerate everyone in their own
intrinsic dimension. A lower power acceleration by orders of magnitude,
but the accelerations are in- different dimensions! I know you have been
reading this (dear Elon) waiting for me to mention big O complexity.
Let's say the small number of competing corporate helper bot dimensions
is roughly 1 dimension, even if it's really 10, or 100.
Given that we're travelling a small ("small") distance along just one
dimension - I say small because improvement has run out, and different
people broooadly get similar results - there is a polynomial
approximation that is very accurate to the phase space volume gained by
a corporation of employees' velocities having received acceleration in
one direction. Anyway I'm hamfistedly trying to argue that the
subscription-chatbot improvement goes as ~ O(n^k).
However, adding dimensions increases complexity (good, in this case) ~
O(exp n).
What this big O stuff means is that for high enough n, 1-dimensional
corporate subscription helper bot land ~ O(n^k), k a big corporate
constant, is always overtaken by
everyone-accelerating-on-their-own-dimension ~ O(exp n). Achievable n is
~ 10000000000, which is pretty big. The volume of the phase space x vs
dx, dx accelerated is titanic compared to x vs dx, dx just accelerated
in one way for everyone (a small number of corporate dimensions). <-
need to develop this theory but I've got a good feeling about it.
The difference in volume of the population phase space in an accelerated
individual-dimensional world rather than the volume of the population
phase space in an accelerated just-a-few-corporate-dimensional world -
well it grows as O ( exp n - m ^ k ), n is large and sm0l, m is corpo &
currently big, k is largeish, n is how many people Lonny personally
liberates from wageslavery, k is how much money Microsoft has vs
the free world.*
What does this titanic difference in volume mean? Well my side's going
to win anyway as n new cyberians grows, I just wrote this phost to
invite my hypothetical good-universe Elon Musk to the best laundromat.
*
n + m ?= 10000000000
**
So by the nature of exp(n) a tiny number of people acting freely -
on the order of 25 - would overwhelm the Microsofts of this world. This
is unobserved, so I think there's a coefficient relating to how much of
one Microsoft each person is. Anyway, the auspices are good.