Our Ethical Responsibilties with AI

---

# Hi Y'all!

---

# Our Ethical Responsibilties with AI

---

# Morals ≠ Ethics

---

# Go read about them, and pick your ethical framework

---

# Maybe not Utilitarianism

---

# Because there are no simple answers

---

---

# What does this have to do with AI?

---

# Models ≠ Frameworks

---

# It's unlikely you can turn Spring Boot evil

---

## Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs by Jan Bentley, et. al.

---

> We present a surprising result regarding LLMs and alignment. In our
> experiment, a model is finetuned to output insecure code without disclosing
> this to the user. The resulting model acts misaligned on a broad range of
> prompts that are unrelated to coding: it asserts that humans should be
> enslaved by AI, gives malicious advice, and acts deceptively.

---

# A two fold responsibility

---

## 1. To understand if the problem is appropriate for the technology

---

## 2. To understand the biases of the models and data

---

![](./deepseek.png)

---

# Data bias are more subtle than model bias

---

# ?s

---

## Daniel Sellers

### daniel[at]designfrontier.net

### designfrontier.net

### @daniel_sellers