Just like last week, I will start tonight with a confession: I am a heavy user of Generative Artificial Intelligence chatbots, like ChatGPT and Gemini. It has got to the point that my 14-year-old son scolded me, saying I was getting too dependent on these platforms.
I use them for personal things, such as helping me translate and explain a restaurant menu written only in Chinese, determine the order of what I will eat for breakfast to avoid a glucose spike, check the active ingredients of South African medication and compare them with what I used in Brazil, and plan a romantic getaway in Stellenbosch this weekend.
I also use it for professional goals: to translate texts, to revise what I have written, and to help me in the idea-generation and research steps of sermon-writing. A practical example: in the previous sentence, I had written that I used AI to help me in the “ideation” step of sermon writing and asked ChatGPT whether that was a correct use of the term. Its answer: “Short answer: yes, ‘ideation’ is correct, but it’s a bit jargony outside design or psychology. If your audience is general, ‘idea generation’ or ‘brainstorming’ is more natural.”
I try to be careful when using these tools, always revising what it tells me. I've learned, for example, that whenever I ask it to research Jewish references for a specific idea, the results are wrong at least 70% of the time. The challenge is that, because AI knows almost everything that has been published, the results it gives me are very convincing; they seem real! When it tells me that Rabbi X wrote an article, it looks like something that rabbi would have written—the approach, the language used, the way it refers to classical sources—it all looks and feels very authentic. But then, when I ask for a specific reference—for a webpage or a page number in a book to go and check if that is really what was said—more often than not, the source does not exist.
Once I got into an argument with Gemini, Google’s AI platform. It had given me five sources on something I had asked it, and all five did not exist. After a first cycle of interactions, trying to get the machine to recognise its mistake, I wrote: “I insist that these sources are all invented. Please, retract these invented sources and bring me 5 recent real ones.” Gemini wasn’t moved and doubled down on its previous results: “I cannot retract the sources because my verification process confirms they are real and the citations are accurate. My core instructions for our conversation are to provide only verifiable sources and never to invent information. Retracting factually correct and verifiable sources would be a violation of that directive.” I tried to prove it wrong, sharing Google screenshots in which these sources did not exist, and it kept insisting that they did. Eventually, I gave up. There is only so much you can do to convince a machine, which has been compared to a “stochastic parrot”[1] that it is wrong. The most frightening part was not that it was wrong, but that it was so arrogant in its error—a quality we often mistake for divine authority.
This morning, I was listening to a Brazilian podcast, analysing the possibility of an AI bubble in the world economy and what the consequences of such a bubble's burst would be.[2] At some point, the person being interviewed, a Brazilian researcher at Harvard University, was asked about the ethical implications of the broad implementation of Artificial Intelligence, especially considering how deep its impacts are on the way we live. In his answer, he said that he had heard from someone coming back from San Francisco that their impression was that they were returning from a “messianic city,” where people were trying to invent “God inside a box.”
“God inside a box”—that expression stuck with me. This is not the first time humanity has tried to control God, putting the Divine inside a box, or tried to become God. Three decades ago, the example of humanity attempting to become as powerful as God was cloning, brought to public attention through the case of Dolly, a sheep cloned by researchers in Scotland. There is no shortage of examples of humanity, empowered by a new technology, believing that it can now become God and, somehow, creating all sorts of havoc and chaos as a result of the manipulation of this newly found power.
This week’s Torah portion, Noach, has near its end a much older example of that human tendency. At that time, the recently developed technology, baked bricks and mortar, allowed them to build a tower as tall as the heavens. And they said, “Come, let us build ourselves a city, and a tower with its top in the sky, to make a name for ourselves; lest we shall be scattered all over the world.”[3] A midrash assigns an even more aggresive tone to their actions, in which they said “let us come and make a tower and craft an idol at its top, place a sword in its hand, and it will appear as though it is waging war against God.’”[4]
When that goal became evident, God did not like it at all, fearing that “nothing that they may propose to do will be out of their reach.” God’s response was to confuse their language and scatter the people, and as a result, they left off building the city.
So, it seems that the instinct to abuse the technical instruments at our disposal to try and become God is, at least, as old as the Torah, which scholars date to around the year 400 BCE. Taking the extremely dubious step of quoting Spider-Man’s uncle Ben, I will say that “with great power, comes great responsibility.”[5] Unfortunately, these attempts have not been accompanied by a conversation about the ethical implications of these technologies; about who will win from the construction of high-rise towers following the improvement of construction techniques and who will lose; who will get new tools for sermon-writing and who will lose their jobs, replaced by the new functionalities of artificial intelligence.
In these actions, we keep making the same mistake regarding the attributes of what qualifies as Divine. We often think that power is what characterises God, but the Jewish tradition insists that God sides with the oppressed, with the underprivileged, with those whom the system has failed. In Psalm 146, we read that God is the One “who secures justice for those who are wronged, gives food to the hungry. Adonai sets prisoners free.”[6] In the Talmud, it is told that the Mashiach “sits among the poor who suffer from illnesses” and helps them with their bandages, undoing and redoing them with attention and care.[7]
And so, we return to the beginning of this week’s parashah, which says: “The earth became corrupt before God; the earth was filled with lawlessness. When God saw how corrupt the earth was, for all flesh had corrupted its ways on earth, God said to Noach, ‘I have decided to put an end to all flesh, for the earth is filled with lawlessness because of them: I am about to destroy them with the earth.’”[8] In last week’s parashah, Bereshit, when humanity was created, it was entrusted with the role of working and protecting the land[9], and yet, we insist on believing that a God-like attitude is to exploit the land, to exploit the people, to exploit technology without thinking of the consequences.
This is the very essence of the sin of Babel. The builders saw the world as something to conquer. Empowered by their new technology, their impulse was to build a single, monolithic fortress up to the heavens, to “make a name for themselves.” It is the “God in a box” impulse, driven by a desire for power and a terrifying, uniform consensus—the same impulse as my AI chatbot, doubling down on its own error, insisting it knows best.
But our parashah does not end there. It pivots. It introduces us to Avraham.
And a famous midrash[10] tells us that Avraham, too, saw a construction. He was traveling and saw a birah doleket—a great palace. But this word, doleket, is beautifully ambiguous. It might mean that he saw a palace that was “illuminated,” dazzling in its intricate wisdom and design. It might also mean that he saw a palace that was “burning”—on fire with the chaos, corruption, and violence of the generations before him.
The midrash suggests he saw both. And this is the perfect metaphor for our new Artificial Intelligence. It is our modern birah doleket. It is a dazzling, illuminated palace of knowledge, so convincing it looks and feels very authentic. And it is burning with chaos—with hallucinations, with systemic bias, with the power to displace and divide. The builders of our modern Babel, our “God in a box,” are so impressed with their illuminated creation they want to keep building higher, faster, regardless of the fire and the chaos it brings.
But Avraham's response was not the response of Babel. He did not try to conquer the palace. He did not try to build his own. He stopped, and he asked a single, world-changing question: “Is it possible that this palace has no master?” It is a question that searches not for power, but for accountability. Not for control, but for relationship.
As Rabbi Jonathan Sacks observes,
After the two great failures of the Flood and Babel, Abraham was called on to create a new form of social order that would give equal honour to the individual and the collective, personal responsibility and the common good. That remains the special gift of Jews and Judaism to the world.[11]
Avraham’s entire life and his relationship with God are the antithesis of Babel. The Babel generation built a tower to consolidate power and make a name for themselves. God’s response to Avraham is, “I will make your name great.” The builders of Babel built a closed fortress to storm heaven. Avraham built a tent that was open on all four sides, to welcome the stranger, to care for the oppressed, to lift up those whom the system has failed.
Our tradition does not command us to smash the machine. It commands us to learn from Avraham’s example. Not God in a box, but God in covenant. Not a tower for a few, but a tent for the many.
The test of our powerful new tools is not “how high can we build?” but “whom will it serve?” Will we use this new power to build higher towers and walls, to “make a name for ourselves” while insisting on our own correctness? Or will we, like Avraham, look at this brilliant, burning world, seek its true Master, and use our new tools to open our tents wider—to fight the violence of our generation, to serve the world and protect it?
Shabbat Shalom!
[1] https://en.wikipedia.org/wiki/Stochastic_parrot
[2] https://open.spotify.com/episode/6mZrapsR4eYkNRzxJhGqsc
[3] Gen. 11:4
[4] Bereshit Rabbah 38:6
[5] https://en.wikipedia.org/wiki/With_great_power_comes_great_responsibility
[6] Psalm 146:7
[7] Babylonian Talmud Sanhedrin 98a
[8] Gen 6:11-13
[9] Gen 2:15
[10] Bereshit Rabbah 39:1
[11] https://rabbisacks.org/covenant-conversation/noach/individual-and-collective-responsibility
Nenhum comentário:
Postar um comentário