We have expressed our concern regarding AI before at some
length. The main concern was that it was not possible to define it, especially
as one tried to legislate controls over it. Unlike the development of what we
now see as the Internet, AI appears to have limited if any benefit to its
users. In fact, one may not even be aware that there is some AI, whatever that
means, intervening in their day-to-day life.
In the early days of the Internet, the user saw the results
and benefits immediately. They could send emails, download and upload files,
get documents in readable form instead of fax. Before web browsers even,
benefits were clear. Modems may have been slow but faxes were slower. Dial up
allowed for users with a telephone line to get results. There was an obvious
and overt progression of benefits. Cable modems allowed increased speed and
browsers allow simplified interfaces. Then the dot com boom exploded with new
applications. It changed the economy.
So far so good. Then came social media. The end of the boom
and the beginning of a propaganda-based paradigm. Anonymous postings did away
with any ability to vet veracity or sources of statements. This was a major
loss. One could “post” or state anything without any ability to verify. Then
add to such apps as Facebook, not just “friend” to “friend” communications but
sidebars of targets promotional propaganda. These apps as we noted decade ago
could psychologically profile a user and them promote in a user specific manner
whatever some third party wanted to get the user to believe. Users could be
manipulated at low cost, real time, and with high targeting efficacy. A deadly
tool. But still visible.
Now comes AI. Whatever it is, it is hidden. Behind a wall of
software. It now knows you; it knows how to manipulate you. It is goal driven
to get you to do something. And you never see it, never get a benefit. It
benefits the supplier solely often causing you harms.
Besides the societal person to person effects of AI, there
are massive ecological effects. For example, the explosive use of electrical
power and generation of heat is just one. The AI farms effectively tax everyone
by the increased demands for power.
How does this power issue work? Simple. Power if delivered
locally and produced separately. For the most part the local power company is
akin to the old local telephone company. It just provides the local lines for
distribution. The actual power is generated by third parties who own and
operate generators. They are like distant telephone companies. There then is
the network agglomerators who interconnect generators to distributors. Here is
where it gets tricky. Periodically the local operators negotiate with the
network agglomerators for power. They agree to buy X at price Y. As the demand
X increases Y increases. This is not linear! A key fact. Thus, when X
increases, Y increases even more so. Namely it gets more expensive per unit!
Who pays, the consumer!
Thus, AI creates a social cost or deficit. It makes us pay
to be manipulated. We see no benefit, we neither get information,
entertainment, or perform efficient transactions. We do get propagandized flows
of persuasion promoting the creators’ intended ideas.
Is this AI, again whatever it is, worth the costs we are now
paying? Purported the AI sucks up massive data, say from medical publications,
and when we ask a question, it forms an answer from this massive collection.
But are the data we collected it from correct? Often it is not. No one vets the
source data. There may a lot of junk. Unreliable “facts” used to create false
conclusions.
Overall, AI is a danger for three major reasons:
First, it is a massive targeted propaganda instrument.
Second, it creates excessive costs to consumers for such
things as energy usage and water supply survival.
Third, it relies upon massive amounts of highly unreliable
input information creating equally unreliable results.
Simply stated, AI, whatever it is, may be not just questionable
but a deadly societal weapon.