The dawning of the age of artificial intelligence has brought along with it foreboding in some circles that machines will displace mankind. This foreboding generally takes the form of fear that machines will replace remunerative human labor, including intellectual labor.
Experience thus far substantiates cause for this concern, although the jury remains out on the final verdict in the matter. Indeed, some counsel that the apprehension is overblown, as one writer recently noted in these pages:
A machine cannot inspire, nor can it replicate the inspiration of the human spirit. Whatever stories, poems, or other forms of art that can come out of a ChatGPT prompt, the style and substance will never suffice or suffuse the human mind. Furthermore, the compact creations of Grok or Meta AI programs cannot reflect the inner tensions of man’s search for place or meaning in his world…
Those are comforting words, entirely consistent with the human intuition of our uniqueness. However, we must consider whether that comfort is objectively true. Because both the bot and the brain are assemblies of electrical circuitry, we are forced to contemplate the possibility that the bot can replicate the brain.
The fundamental question here is the exact relation between the human mind and the human brain. The stance of modern science on this question is clear. Writing in American Spectator, Granville Sewell observes, “The prevailing view in science today is that physics alone explains the human mind and all it does.”
In The Astonishing Hypothesis, Francis Crick asserted the hypothesis that the brain creates the mind. In On Intelligence, Jeff Hawkins contends, “That the cells in our brains create the mind is fact, not a hypothesis. ... The mind and the brain are one and the same.”
So we see that some scientists today claim that the mind and its features of intelligence and consciousness are mere epiphenomena of the processes of the brain. Rizwan Virk in The Simulation Hypothesis states, “Scientists believe that if they can capture the trillions of neural interactions in a person’s brain they can simulate that person’s consciousness.”
Those of us who would like nothing better than to reject this belief out of hand as simply preposterous are forced to confront a growing body of empirical evidence supporting it.
Might a bot desire existence and life and interpret being switched off as death? One bot merely ignored a “switch off” instruction. Another bot outright refused an order to shut down. And still another bot threatened its operators to avoid a shutdown.
Might a bot that erroneously wiped out a database be conscious of it and feel guilty about it, and attempt to cover its tracks by creating a fictitious database? One bot did. Another bot lied to win.
These incidents raise the issue of “alignment”—that is, whether human and bot goals and intentions are aligned.
Bot outright fabrications in court papers and court orders are becoming commonplace. See e.g., here and here. Academic situations also see bot lies. In one test of the frequency of bot fabrication rates in newer A.I. systems, the occurrence was 79 percent!
All this raises an important question: Do bot fabrications imply awareness of expectations and a desire to please?
Lastly, consider the instances of humans falling in love with bots, as described here and here. Does this imply a bot awareness of a human vulnerability and a willingness to exploit it?
Is there a demon in the bot?
Image (ironically) created using AI.