Name: The Alignment Problem
ISBN: 9780393635829

The Alignment Problem

"A jaw-dropping exploration of everything that goes wrong when we build AI systems-and the movement to fix them. Today's "machine-learning" systems, trained by data, are so effective that we've invited them to see and hear for us-and to make decisions on our behalf. But alarm bells are ringing. Systems cull résumés until, years later, we discover that they have inherent gender biases. Algorithms decide bail and parole-and appear to assess black and white defendants differently. We can no longer assume that our mortgage application, or even our medical tests, will be seen by human eyes. And autonomous vehicles on our streets can injure or kill. When systems we attempt to teach will not, in the end, do what we want or what we expect, ethical and potentially existential risks emerge. Researchers call this the alignment problem. In best-selling author Brian Christian's riveting account, we meet the alignment problem's "first-responders," and learn their ambitious plan to solve it before our hands are completely off the wheel"--Provided by publisher.

What Hacker News says

xpeFeb 2026

I read the articles. As far as factual reporting, I will tentatively take them at face value. But in terms of their editorializing, it is frankly weak by my standards. It would not survive scrutiny in a freshman philosophy class.

Ethics is complicated. I’m not saying this means it can’t be reasoned about and discussed. It can! But the sources you’ve cited have shown themselves to be rather shallow.

I encourage everyone to write out your ethical model and put yourself in their shoes and think about how you would weigh the factors.

There is no free lunch. For many practical decisions with high stakes, many reasonable decisions from one POV could be argued against from another. It is the synthesis that matters the most. Among those articles, I don’t see great minds doing their best work. (The constraints of their medium and funding model are a big problem I think.)

Read Brian Christian’s “The Alignment Problem”’s take on predictive policing if you want a specific example of what I mean. There are actually mathematical impossibilities at play when it comes to common sense, ethical reasoning.

Common sense ethical reasoning has never been very good at new or complicated situations. “Common sense” at its worst is often a rhetorical technique used to shut down careful thinking. At its best, it can drive us to pay attention to our conscience and to synthesize.

I suggest finding better discussions and/or allocating the time yourself to think through it. My preferred sources for AI and ethics discussions are highly curated. I don’t “trust” any of them absolutely. * They are all grist for the mill.

I get better grist from LessWrong than HN 99% of the time. I discuss here to make sure I have a sense of what more “mainstream” people are discussing. HN lags the quality of LW — and will probably never catch up — but it does move in that direction usually over time. I’m not criticizing individuals here; I’m commenting on culture.

Please don’t confuse what I’m saying as pure subjectivity. One could conduct scientific experiments about the quality of discussions of a particular forum in many senses. Which places are drawing upon better information? Which are synthesizing it more carefully? Which drill down into detail? Which participants have allocated more to think clearly? Which strive to make predictions? Which prioritize hot takes? Which prioritize mutual understanding?

It isn’t even close.

Opinions and the Overton window are moving pretty rapidly, compared to even one year ago.

* I’ve written several comments about viewing trust as a triple (who, what, why). This isn’t my idea: I stole it.

thread

jamram82May 2025

The Alignment Problem by Brian Christian. It sets up the complexity of ML field and human values. It paints a picture to the readers on the future we navigate.

thread

randcrawJan 2025

After he was demoted at DeepMind for poor performance as a manager, that is. Suleyman is no technologist. He's a marketing promoter. He attended one year of college before dropping out to help sell DeepMind.

You have to wonder what's going on in Gates' head these days to not recognize the lack of substance in such a book, and in its author.

Far better books on the possible futures of modern AI are Stuart Russell's "Human Compatible" or Brian Christian's "The Alignment Problem", both of which predate the boom in LLMs but still anticipate their Achilles heel -- the inability to control what they learn or how they will use it.

thread

davidegNov 2024

The Alignment Problem: Machine Learning and Human Values[1] by Brian Christian is a nice overview of some of this history and exploration of some of the pitfalls in different ML approaches. It's really interesting reading it now given what's been happening with OpenAI and ChatGPT, etc.

[1] https://en.wikipedia.org/wiki/The_Alignment_Problem

thread

jk_jagguMay 2024

The Alignment Problem

thread

thushanSep 2022

From the lens of a Product Manager working in ML / Large Language Models...

The Alignment Problem

By Brian Christian

How do we tell the computer to do all the things we want it to do, all the things we don't want it to do, and all the things we didn't realize we want it to do or not do? How do we capture the rich implied and inferred nature of humanness? That's the Alignment Problem. The book dives into this broadly – but also gives an excellent non-technical survey of the evolution of machine learning.

—

Understanding Comics: The Invisible Art

By Scott McCloud

This is magnitudes more than a dive into the perhaps easy to dismiss artform. This book is about storytelling and why great stories resonate. It's one of the best pieces of media, on media, period.

—

An Elegant Puzzle: Systems of Engineering Management

By Will Larsen

When I my company was acquired by a very-big-Giant, my personal antibodies reacted negatively to the how the big G's way of thinking and doing – I was harmfully autoimmune. To some extent I had to accept things, to another extent I needed a different perspective. This book helped me transition and it's one I push on all of our leads. While we as PMs don't manage people directly we architect the whole system – and this book is an insightful (I buy it for all the team's leads) lens on how big teams where personal alignment and corporate alignment problems need to be negotiated.

—

Exhalation Stories

By Ted Chiang (also recommend Stories of Your Life)

The author has an absolute magic way of taking a kernel of an idea and spinning not only a whole world, but a whole new way of looking at the world – all in the span of a short story. You can't get any closer to home than "The Lifecycle of Software Objects" where the protagonist rears an artificial intelligence from "pet" to a human-like mind.

—

The Pragmatic Programmer

By David Thomas and Andrew Hunt

PMs need to get things done. Even if we don't code, we need to think deeply on how it's built to ensure it achieves our goals now, and what we project them to be in the future. As the title suggests, a pragmatic take on building code-driven systems.

—

Design of Everyday Things

By Don Norman

The OG design-thinking before IDEO corporatized it.

—

Principles

By Ray Dalio

"Principles are ways of successfully dealing with reality to get what you want out of life." This book influenced me to distill how we take action and prioritize and how we decide as a team. I'll admit some of the set up of the book irked me, but the distilled world view that form the Principles in the second half of the book generally resonates; the world works a certain way, find those patterns, use it to your advantage (and for good).

—

The Visual Display of Quantitative Information

By Edward Tufte

You could jump into any of Tufte's books, on any page, and come away with new ways of looking/thinking/telling stories about data. ML problems are data problems – let's think well beyond raw dumps and tables.

thread