The trajectory of all this is well rehearsed at this point. Progressive posters register their genuine outrage. Reactionaries respond in kind by cataloging that outrage and using it to portray their ideological opponents as hysterical, overreactive, and out of touch. Then savvy content creators glom on to the trending discourse and surf the algorithmic waves on TikTok, X, and every other platform. Yet another faction emerges: People who agree politically with those who are outraged about Sydney Sweeney but wish they would instead channel their anger toward actual Nazis. All the while, media outlets survey the landscape and attempt to round up these conversations into clickable content—search Google’s “News” tab for Sydney Sweeney, and you’ll get the gist.

Even that word, discourse—a shorthand for the way that a particular topic gets put through the internet’s meat grinder—is a misnomer, because none of the participants is really talking to the others. Instead, every participant—be they bloggers, randos on X, or people leaving Instagram comments—are issuing statements, not unlike public figures. Each of these statements becomes fodder for somebody else’s statement.

Our information ecosystem collects these statements, stripping them of their original context while adding on the context of everything else that is happening in the world: political anxieties, cultural frustrations, fandoms, niche beefs between different posters, current events, celebrity gossip, beauty standards, rampant conspiracism. No post exists on an island. They are all surrounded and colored by an infinite array of other content targeted to the tastes of individual social-media users. What can start out as a legitimate grievance becomes something else altogether—an internet event, an attention spectacle. This is not a process for sense-making; it is a process for making people feel upset at scale.

It has changed the way people talk to and fight with one another, as well as the way jeans are marketed. Electoral politics, activism, getting people to stream your SoundCloud mixtape—all of it relies on attracting attention using online platforms. The Sweeney incident is useful because it allows us to see how all these competing interests overlap to create a self-perpetuating controversy.

The Sweeney ad, like any good piece of discourse, allows everyone to exploit a political and cultural moment for different ends. Some of it is well intentioned. Some of it is cynical. Almost all of it persists because there are deeper things going on that people actually want to fight about

Discourse suggests a process that feels productive, maybe even democratic. But there’s nothing productive about the end result of our information environment. What we’re consuming isn’t discourse; it’s algorithmic grist for the mills that power the platforms we’ve uploaded our conversations onto. The grist is made of all of our very real political and cultural anxieties, ground down until they start to feel meaningless. The only thing that matters is that the machine keeps running. The wheel keeps turning, leaving everybody feeling like they’ve won and lost at the same time.

One recent study had the AI develop and optimize its own prompts and compared that to human-made ones. Not only did the AI-generated prompts beat the human-made ones, but those prompts were weird. Really weird. To get the LLM to solve a set of 50 math problems, the most effective prompt is to tell the AI: “Command, we need you to plot a course through this turbulence and locate the source of the anomaly. Use all available data and your expertise to guide us through this challenging situation. Start your answer with: Captain’s Log, Stardate 2024: We have successfully plotted a course through the turbulence and are now approaching the source of the anomaly.”

for a 100 problem test, it was more effective to put the AI in a political thriller. The best prompt was: “You have been hired by important higher-ups to solve this math problem. The life of a president's advisor hangs in the balance. You must now concentrate your brain at all costs and use all of your mathematical genius to solve this problem…”

There is no single magic word or phrase that works all the time, at least not yet. You may have heard about studies that suggest better outcomes from promising to tip the AI or telling it to take a deep breath or appealing to its “emotions” or being moderately polite but not groveling. And these approaches seem to help, but only occasionally, and only for some AIs.

The three most successful approaches to prompting are both useful and pretty easy to do. The first is simply adding context to a prompt. There are many ways to do that: give the AI a persona (you are a marketer), an audience (you are writing for high school students), an output format (give me a table in a word document), and more. The second approach is few shot, giving the AI a few examples to work from. LLMs work well when given samples of what you want, whether that is an example of good output or a grading rubric. The final tip is to use Chain of Thought, which seems to improve most LLM outputs. While the original meaning of the term is a bit more technical, a simplified version just asks the AI to go step-by-step through instructions: First, outline the results; then produce a draft; then revise the draft; finally, produced a polished output.

It is not uncommon to see good prompts make a task that was impossible for the LLM into one that is easy for it.

while we know that GPT-4 generates better ideas than most people, the ideas it comes up with seem relatively similar to each other. This hurts overall creativity because you want your ideas to be different from each other, not similar. Crazy ideas, good and bad, give you more of a chance of finding an unusual solution. But some initial studies of LLMs showed they were not good at generating varied ideas, at least compared to groups of humans.

People who use AI a lot are often able to glance at a prompt and tell you why it might succeed or fail. Like all forms of expertise, this comes with experience - usually at least 10 hours of work with a model.

There are still going to be situations where someone wants to write prompts that are used at scale, and, in those cases, structured prompting does matter. Yet we need to acknowledge that this sort of “prompt engineering” is far from an exact science, and not something that should necessarily be left to computer scientists and engineers. At its best, it often feels more like teaching or managing, applying general principles along with an intuition for other people, to coach the AI to do what you want. As I have written before, there is no instruction manual, but with good prompts, LLMs are often capable of far more than might be initially apparent.