返回资讯中心
外部精选
软件工程
#开发者实时流

Warped Mirrors

AI discourse is so incredibly confusing. As someone who works in tech, has been writing about AI for over a year on this blog, and attempts to stay up-to-date with all the latest industry news, I cannot imagine what it …

Hacker News 最新jep8882026/10/16 · 2026/10/1612 分钟阅读

以下正文同步自 Hacker News 最新,版权归原站所有,已转换为易读排版。

James Pember

Sep 08, 2026

1

Share

AI discourse is so incredibly confusing. As someone who works in tech, has been writing about AI for over a year on this blog, and attempts to stay up-to-date with all the latest industry news, I cannot imagine what it must feel like for someone trying to follow along from the outside.

Just look at this collection of recent articles from The Guardian over the past few days. All the talking points are there! AI job loss, climate concerns and outraged artists mixed in with medical, academic and scientific breakthroughs.

What’s incredibly confusing about AI, and actually every tectonic technology shift, is that it’s impossible to draw simple, binary conclusions about these new technologies.

This is because tools aren’t neutral. Bill Gates wrote the other day on his blog that “AI will either be the greatest equalizer ever invented, or the worst source of injustice”, but I believe the truth is that there will be no “or” in between “greatest equalizer” and “source of injustice”. It’s going to be both, all at the same time. AI companies proselytising one way or the other is ultimately just marketing.

Increasingly I believe that AI is the ultimate Rorschach-test. AI is a mirror, and when we stare into it, we see what we want to see reflected back at us. For some, it’s climate disaster and economic destruction, for others, a golden age of science and medicine. AI becomes just another surface to project our existing beliefs onto.

Antoine Le Gonidec’s “resignation” from Debian

When we stare into that mirror, we see what we want to see, and the warped reflections that we meet reveal that we are trapped between two lazy extremes:

The naive delusion that history is littered with moral panics (skateboarding, “the idiot box”, heavy metal etc) and we should ignore them

The narcissistic delusion that our moment is entirely unique (i.e. “we are living through unprecedented times”)

Moral panics aren’t new, but are they justified?

On the first point, I’m actually fairly cautious about dismissing past moral panics. While some were genuinely ridiculous, the passing of time makes other topics far murkier. Resolving the cost/benefit of social media, online pornography or online gaming is far more complex.

At the same time, I’m equally cautious of presentism, the idea that as humans we have a horrible bias to believe we are living in “unprecedented times”, which we of course, are not. Humans are remarkably adaptable, and we’ve lived, and thrived through many a technological upheaval before.

“If you believe that what’s happening is something completely new, something that has never happened before, and that history offers no guidance for it, then historical guidance looks irrelevant, and the people offering it look ignorant or dishonest. You end up, in effect, refusing to consult the only evidence base you have about what actually works. Every response to every past crisis, however relevant, gets filtered out as inadequate to the unprecedented present, and you end up making it up from scratch, badly, with tremendous urgency and very limited information.” - JA Westenberg

If we are to avoid this ingrained biological bias we have, and try to understand AI as part of a wider arc of history, the story is ultimately a positive one!

The internet, social media and globalisation have caused real social dilemmas, but the overarching theme of the last 200 years is still one of progress and prosperity. What’s that quote that kicks around online from time to time? “Even kings didn’t have warm showers 200 years ago". A little trite perhaps, but worth keeping in mind.

That said, electricity, factories, the printing press, and even the internet didn’t naturally become complicated-yet-ultimately-net-positive tools on their own. We (society) made them net-positives through (often difficult) social negotiation. Industrialisation was a disaster for the working class (just go read Oliver Twist..) until we invented child labor laws, unions, and the 2-day weekend. The printing press brought a century of chaotic religious warfare until society constructed new social infrastructure like peer-reviewed scientific journals, libel laws, and journalistic ethics.

If AI is to push us upward and along the same path of generally improving living standards that started with the Industrial Revolution, new “things” need to be built, and I don’t just mean technology. New policy, laws, norms and values need to be debated, evaluated and established. We need frameworks and guardrails to ensure that whilst we reap the tremendous productivity benefits AI promises, we find ways to encourage and enjoy human art, stay present in the physical world around us and steer clear of the many dystopian scenarios we’re constantly served online.

正文由 探How 从来源站点 RSS 同步,内容未经改写;遇到排版缺失或需要图片、视频时请以原文为准。

打开原文
去社区讨论 →