Brian Roberg's Blog
Thoughts on ideas concerning the church, society, and technology.
-
Babel and Epistemological Crisis
Read moreLLMs are really good at translating between human languages. You can connect your smartphone to an LLM and use it to translate a conversation in real time. Combine that capability with earbuds and you have a real-life version of the babelfish from Hitchhiker's Guide to the Galaxy.
Douglas Adams' "babelfish" was, of course, a reference to Genesis 11. In that passage God confuses human language in judgment (and mercy) on mankind for their self-satisfied hubris in building the Tower of Babel. This account is one the Bible's most significant comments about how God views human technology. The fact that we've now produced technology that seems to undo the judgment at Babel does not seem to bode well for our future.
-
What Is an AI Agent?
Read moreIf you want to understand important AI concepts, you'll want to get familiar with AI "agents." In the fast-moving world of AI, this particular term has had quite the ride. At this time last year, "agent" was mostly a buzzword: lots of promise, but not yet much substance. But by the end of 2025, agents were delivering on all the hype, and then some.
What is an AI agent? A good place to start is the familiar meaning of "agent": someone who acts on behalf of someone else. A traditional chatbot like ChatGPT is not an agent because it can't do anything other than write text in response to your prompts. But if you give an LLM access to a set of tools and instructions about how to use them, you can give the model the capacity to act—to do something on your behalf.
-
Book Review: Made in Our Image by Stephen Driscoll
Read moreArtificial intelligence is bringing changes quickly, and people have lots of opinions. It's not the easiest topic to get a handle on. If you're looking for an accessible book to help you see AI through a biblical lens, I recommend Made in Our Image by Stephen Driscoll.
There are several beneficial features of this book:
- It's biblical: Driscoll's primary concern is to help readers bring biblical principles to bear on AI.
- It's accessible: He doesn't presuppose a lot of technical knowledge, nor familiarity with other writing about the impact of technology on society. It's also not super-long (192 pages).
- It's reasonably timeless: Because his purpose is to apply biblical principle rather than describe the technology itself, what Driscoll says about AI should remain relevant for at least a few years (which is a very long time in the AI world).
- It's opinionated: he takes definite positions regarding some key questions about AI.
-
Conversation Habits, Followup
Read moreA followup to my post the other day about conversation habits with LLMs: Anthropic just added the following to the system prompt for their Claude models:
"If the person is unnecessarily rude, mean, or insulting to Claude, Claude doesn't need to apologize and can insist on kindness and dignity from the person it’s talking with. Even if someone is frustrated or unhappy, Claude is deserving of respectful engagement."
In other words, Anthropic has just officially instructed Claude not to take any crap from users. (H/T Simon Willison)
-
Conversation Habits
Read moreEven if you know that an LLM is not a person, conversing with a chatbot is still verbal communication. Partially because I know that an LLM is not a person, I don't communicate the same way I do with a person. But sometimes I wonder whether the habits I might establish in communication with LLMs will inadvertently spill over into my habits of communications with actual people.
-
Pitch Tipping and Awareness of AI in Baseball
Read moreSomeday we're going to look back on the 2025 Major League Baseball (MLB) season as the first one in which AI began to visibly affect how the game is played on the field. (It's been influential off the field as a tool for analytics for years.) As of now, though, very few people seem to have noticed.
I saw it in two ways, one of which has gotten a lot more attention than the other.
If you followed baseball this year, you may have noticed there was a lot of talk about pitch tipping. (For the uninitiated: a pitcher "tips" his pitches when he inadvertently gives subtle clues to the other team about which pitch he's about to throw.) In past years, there might be a few stories a year about pitch tipping. This year, it seemed like there was a new story every week. The difference was notable enough that The Athletic wrote an article about it (paywall).
That story examined many angles on the phenomenon, but did not mention AI. Now if pitch tipping were the only unusual event suggestive of AI's influence this year, I probably wouldn't write about it. (Pitch-tipping is a phenomenon that tends to attract attention through simple hysteria, sometimes without any actual pitch-tipping occurring at all.) But there's another data point about a different part of the game that makes me think that there's a pattern.
-
How AI Broke Aristotle
Read moreIt's hard to describe abstract ideas like the fragility of society's understanding of personhood. But in an attempt to do so, I'd like to invite you on a thought experiment with me.
Imagine you had a time machine and could go back in time to another era of human history. In this case, you need to go all the way back to… 2017. That's far enough to reach a time before the first large language models (LLMs) were introduced.
Having traveled back to that far-distant time, now imagine yourself walking around your neighborhood. As you walk, you happen upon a snippet of writing. It could be on a slip of paper on the ground, it could be a random tweet on your phone, it could be any medium at all. Even without knowing any context or background of the writing you found, if that writing contains a coherent thought expressed in language then you know something incredibly significant: it was written by a person.
-
How Meta Is Pitching Their New AI Glasses
Read moreMeta announced a new model of "AI glasses" this week. While this is interesting on a technological level, what really struck me was the angle they chose as their selling point.
“Meta Ray-Ban Display glasses are designed to help you look up and stay present.”
That was the second sentence of the announcement, preceded only by the lede sentence that stated the fact of the product's debut. Meta's sales pitch, then, is the promise of access to smartphone-like capabilities without it feeling (or looking) like you're using a device.
In other words, Meta is introducing a device that provides comprehensive mediation of the user's interactions with the world, while promising that those interactions will feel less mediated. This is precisely the kind of re-definition that can make a new medium profoundly disruptive.
-
Craving Garden of Eden Intimacy
Read moreThis is a guest post by my friend and co-worker Jason Maas.
After creating the entire universe and planet Earth, God created a special home to share with his image bearers. “The Lord God planted a garden in Eden, in the east, and there he placed the man he had formed.” (Genesis 2:8) In the garden of Eden God walked and talked with the first humans that He had created in his image. Can you imagine what that was like for Adam and Eve? God, who is all-knowing, always available, and lovingly kind to the core, was right there, directly communicating with all of the human inhabitants of the universe.
When Adam and Eve disobeyed God and sinned one of the worst consequences was a break in this special access and relationship with God. “So the Lord God sent him away from the garden of Eden to work the ground from which he was taken. He drove the man out and stationed the cherubim and the flaming, whirling sword east of the garden of Eden to guard the way to the tree of life.” (Genesis 3:23-24)
What a tragic loss! In this life, on this Earth, the rest of us will never know what it was like to have the kind of access to God that Adam and Eve had in the garden of Eden. Until now, says the cunning serpent-like world of chatbot generative AI.
-
The Power of LLM Context
Read moreToday I learned how much leverage you can gain by using a tool that automates feeding context to an LLM.