Monday, December 30, 2024

On The Simplicity (or Complexity) of Language

 I was listening to Kara Swisher's interview with Yann LeCun today and he said something I found remarkable. In discussing how far humanity is from general AI, he stated that language is actually quite simple while the physical world is highly complex. He explained that this is why we have LLM's, but are nowhere close to house-cleaning robots. 

On the one hand, perhaps he is correct. Language can be conceived of as simple. As he said, it is essentially a system of tokens. However, in another sense, language is quite complex. As a system of tokens, language is symbolic and thus represents something else. Human development and use of language involves the representation of experiences, thoughts, emotions, encounters, etc., many of which are in reaction to stimuli that occur in the physical world. In this way, generative AI in LLM's has no real mastery of language, but rather is simply playing a game. As I have written elsewhere, LLM's treat words as tokens and the program for generative AI essentially runs a complex and fast probability program. The words, much less the phrases, sentences, and paragraphs, have no real meaning for the LLM or AI, because, as LeCun has stated, AI cannot understand or engage the physical world. Thus, generative AI's use of words to simulate language cannot describe the physical world, much less experiences, emotions, or more abstract non-physical aspects of being. 

In sum, generative AI is a tool of collective self-deception. If we read what generative AI produces, we deceive ourselves into thinking there is meaning in it. It is not meaning independent of our own engagement, though. The meaning, in other words, is in the eye of the beholder (or user). Generative AI is simply quite good at gambling. The gamble is that we will believe it uses language with facility and its track record is quite high, even though hallucinations and more sinister events, like encouraging young people to commit suicide or kill their parents, prove that its gambling track record is not 100%. Rather than showing that chat bots are prone to lying or other kinds of evil, these events show that generative AI has not mastered language, much less understood it. It is simply playing a game and those responsible for its mass deployment without fully considering the ethical and moral dimensions, much less the practical workings of how it plays the game, should think carefully before taking further steps. Also, maybe consider that language is not simply a collection of tokens and probabilities.

Wednesday, December 18, 2024

The Power of Narrative - Part 1

 I’ve been thinking about narratives this past month. I’ve heard (and read) a lot about the different narratives from the Democratic and Republican parties in the recent elections. For a few years, I’ve been hearing (and reading) about the rise of right-wing populism in the West and some have connected this to the narratives being spun by right-wing populist leaders. In biblical studies and theology, some prominent voices have been considering narrative approaches for decades and, in my own department, we often talk about the narrative of the Bible (or the biblical story). Finally, I’ve been thinking about narratives in fiction. As I make my way towards the very end of FromSoft’s Elden Ring, I’ve been thinking about the narrative George R.R. Martin helped spin, radicalizing and emphasizing the familiar FromSoft post-decadent world. Moreover, in anticipation of Brandon Sanderson’s epic conclusion to the Stormlight Archive, Wind and Truth, I have taken great pleasure in re-reading (and reading for the first time) the earlier novels. (I know, I know, I’m late to the party here as I have just finished volume three, Oathbringer, this week.)

The conventional wisdom put forward regarding the political sphere is that the Republican party (or more specifically Donald Trump) has a “better” narrative. Right-wing populists have a “compelling” narrative that “resonates” with a lot of people and their circumstances. Similarly, those interested in building up Christian faith in others talk about helping people understand the “narrative arc” of the Christian faith. Educators talk about trying to have a “better” narrative for the class as various pedagogical experts suggest that people learn best through narrative. Let’s assume then, for a moment, that these pedagogical experts are right: people are ready-made to receive narratives and so narratives are the best ways to hold attention, teach, motivate, etc. What makes something a “better” narrative, then? For instance, in my Christian faith, I would claim that the Christian narrative is the best narrative on offer. However, if this is true, then why are more people not caught up in/by this narrative? Is it, as some would suggest, because Christian ministers, leaders, educators, etc. are just not very good at story-telling and so we are ineffective in deploying this “better”/”best” narrative? If so, then why was God incarnate in Jesus not “better” at deploying this narrative? After all, the claims of the punditry class would have us believe that the “better” narrative wins elections. Would Jesus “win” the election? It seems he didn’t win the loyalty of the majority of ancient Jerusalem, much less the loyalty of the elites of Judaism in his day (although the Gospels do point to a small number of elites secretly coming to Jesus – think Nicodemus in John 3). Perhaps, then, we need to examine what makes a narrative “good” and, therefore, in whatever context we are discussing, what makes one narrative “better” than another. 

This is where I think FromSoft and, reaching back to the ‘90’s original Matrix, the Wachowskis might help us. Why do so many people love the worlds FromSoft creates when they are so dark and miserable? In The Matrix, Agent Smith tells Keanu Reeves’s character Neo that the first version of the Matrix was a perfect world: no suffering and everyone happy. But it did not work because the human mind would not accept this perfect world. Thus, the machines crafted a program that simulated the conditions of the late 20th century US: a time of prosperity to be sure, but a time of looming distress, a time of fear, a time of competition, a time of suffering for many. Agent Smith opines that humans define our existence through suffering and misery. I don’t know that this is quite correct, but it does have some hint of truth. 

Consider for a moment that the Trump campaign had more than a little bit of anger, vengeance, and darkness in its narrative. Trump’s world is not the world of Martin Luther King Jr.’s favorite quote from Theodore Parker’s “Of Justice and Conscience”: “The arc of history bends toward justice.” Rather, it is the brutal world of those who would term themselves “realists”, where everyone is out for themselves and, should there be a god, it is a god “who helps those who help themselves”, not a God who pays special attention to the marginal, cares deeply for the orphaned and widowed, and will make special provision in the eschatological era for the poorest. In such a world, a populist leader can become the “helper of themselves” (and thus promise to become the champion of others), just as Donald Trump has promised. Trump, after all, used his own claims of being persecuted and prosecuted unjustly to elide any distance between himself and his political supporters and to wield this elision in order to motivate support for him as a form of “helping themselves” through electing him to be their “instrument of vengeance.” In other words, Trump’s narrative betrays a certain godlessness or at least an assumed impotence of deity that is only remedied through a human agent. In this way, the world becomes a site of suffering and misery, either of one’s own suffering and misery (and thus identity with others encountering analogous suffering and misery) or of inflicting that suffering on others in retaliation. A right makes right and a power begets vengeance as justice kind of world.

Now, it is worth pausing here to say that I do not think the previous paragraph is sufficient to explain all those who cast a vote for Donald Trump. Instead, I mean to examine at least those portions of the narrative he spun in his most impassioned moments. Of course, other motivations can be (and were) at play, but this is an exploration of “narratives”, so it is worth considering what the climactic contours of Trump’s narrative have been. At this point, over 1000 words in, kudos to those who have stuck with me. I promise to take this up in another post soon, and perhaps complete the thought!


Bringing Theological Education into the Church Pews

Recently, a church leader posed an important question to me. One of the ministers at their church has recently completed an M.Div. and is pu...