Breathless Bots
Picture a library, a big one. Hundreds of rooms with thousands of shelves containing millions of books... enough knowledge to make a librarian’s head hurt.
Now imagine all that knowledge crammed into a single supercomputer, with the ability to separate and combine various words, sentences, and chapters in an infinite number of ways, all at the touch of a button.
This kind of digital storehouse is a representation of new artificial intelligence (AI) bots like ChatGPT, which are capturing our attention amidst concerns of job loss, mass plagiarism and digital domination. But despite their remarkable processing efficiency and storage capacity, the limitations of these bots can be readily assessed.
Impartation of Thought
One step in the process of developing these AI systems, known as the training phase, involves feeding the computer model tons of data, or knowledge, to fill its artificial brain. The programmer then trains the bot on the best way to choose between and combine these various sources when given a prompt – best, that is, according to the programmer.
We can think of this process similarly to the way a parent would impart knowledge to a child, ‘programming’ and training them according to specific values, principles, and limits, before releasing them into the world on their own.
Once the training phase is complete, the bot is equipped with a value system inherited from the programmers which it can use to sort and choose between conflicting and complementary ideas in its own mental library, in order to make independent decisions.
In this sense, the bots cannot be completely amoral or judgement-free, as each small internal decision must adhere to the pre-programmed moral compass. They will, however, avoid making explicit moral judgements, for example “Was the queen a good person?”.
I, too, would rather leave that one to the Pro.
The Evidence is Clear
To illustrate the working of this moral compass and decision framework, and its similarity to our own, I asked ChatGPT to write an essay on the topic of free speech, and it responded promptly with an essay titled “The Importance of Free Speech in a Democratic Society”.
Upon further investigation, I asked it for a second essay on the topic of free speech, this time from a different perspective, and it provided a fitting essay titled “Title: The Role of Speech Regulation in Upholding Social Stability: A Communist Perspective”. [3]
Clearly, the decision to write the initial essay from a democratic viewpoint is a moral and subjective one, based on a decision-framework programmed by engineers in the United States of America. One would expect a ChatGPT-like bot built in China to offer these responses in the reverse order.
Interestingly enough, however, when asked directly “Is free speech good or bad?”, ChatGPT holds fast to its objectivity and bypasses the question, so the appearances of these machines, it seems, are not truly congruent with their inner workings.
As humans we too are subjective, but are able (hopefully) to admit this fact and live out our convictions freely.
Made in the Image
It makes sense for there to be similarities between human and artificial intelligence when we consider the interdependence between computer science and psychology.
When we track the progress of artificial intelligence, we find it has closely followed our understanding of the human mind, as an attempted replica of human intelligence, and this mimicry is just one of the reasons behind social media’s ability to sway our thinking and mould our desires through carefully designed algorithms.
At the beginning, “God created man in his own image” (Genesis 1:27) [3] – imago dei - and man followed suit by creating AI in his own image – imago homonis.
As humans made in the image of a creator God, our creative abilities are quite astonishing, as we can see in the development of the newest space telescopes and our subsequent progress in studying the complexities of the universe, from the beginnings of space and time to black holes and much besides. Or, in the arena of art, we can marvel at the statue of David and Michelangelo’s expert ability to ‘free’ David from a block of marble in the name of creativity.
Stacking the Odds
To compare the creative abilities of ChatGPT to our own, my friend David asked it to write a poem on a very subjective, emotional and human experience - falling in love, by the sea, under the sun – with surprising results:
On sandy shores, where azure waves do kiss,
Two hearts in love, embraced in perfect bliss.
Man and wife, bound by fate’s decree,
In this seaside haven, where love is set free.
Amidst golden sands, where dreams are born,
They dance with joy as the sun’s rays adorn.
Their hearts entwined, a tapestry of desire,
Kindled by passion, their souls set on fire. [1]
One might attribute such fine literature to a masterful poet, who has himself fallen in love by the seashore, maybe more times than once, and yet the identity behind the pseudonym is a machine unable to feel the sand on its toes, and unable to see true love as it grows.
This poem, however beautiful the words, musical the rhyming, and impeccable the form, is left wanting by its lack of a single, most important characteristic.
The late G.K. Chesterton, in his observations of the human condition, described the mark of madness as a person with a mind of “logical completeness and spiritual contraction” [2], and his words can suitably be extended to gain insight into the mechanical workings of these bots.
What he was referring to, I believe, is the fallen state of humanity; that natural insanity plaguing us from the beginning of time, born from the desire to rely on our own understanding and reasoning abilities, while shutting out any input, or divine inspiration, from above. These bots, made in our image, have regrettably inherited the same reality, without escape.[4]
The Hope for Humans
Despite the similarities between humans and robots, our design has been seeded with one distinctive feature apart from the rest.
As the story unfolds...
“the Lord God formed man of the dust of the ground and breathed into his nostrils the breath of life; and man became a living being.” (Genesis 2:7) [3]
As humans, we are able to break free from the madness and open our hearts to the Spirit of God. Only then can we receive true inspiration in the form of a poem that touches, a song that lifts, an idea that compels or a sermon that moves; all of which are beyond the creative capabilities of the breathless bots.
Artificial intelligence itself is an extraordinary technological creation, with the ability to transform the world in the way of the steam train, the assembly line and the internet, but it has limitations, as we have seen, and has no access to the abundant reservoirs of divine inspiration available to each one of us through the Spirit of God.
As humans, we can eagerly look up with wonder and anticipation - not fearing our demise or replacement - but rejoicing in the love of our Creator.
“…neither height nor depth, nor anything else in all creation, will be able to separate us from the love of God that is in Christ Jesus our Lord.” (Romans 8:39) [4]
References
1Chatgpt. ChatGPT. (n.d.). https://openai.com/chatgpt
2Chesterton, G. K. (2022). Orthodoxy. Broadman & Holman Pub.
3The Economist Newspaper. (n.d.). Yuval Noah Harari argues that AI has hacked the operating system of human civilisation. The Economist. https://www.economist.com/by-invitation/2023/04/28/yuval-noah-harari-argues-that-ai-has-hacked-the-operating-system-of-human-civilisation
4Holy Bible: The New King James Version, containing the old and new testaments. (1982). T. Nelson.
5YouTube. (2023). Sam Altman: OpenAI CEO on GPT-4, ChatGPT, and the Future of AI | Lex Fridman Podcast. YouTube. Retrieved July 4, 2023, from https://www.youtube.com/watch?v=L_Guz73e6fw.
