Daniel Miessler

There's a lot of conversation about how AI could go bad at an existential level.

One of the most serious problems with AI is the potential to become overreliant on it.

The metric for AI is whether or not it’s improving the lives of conscious creatures.

There are lots of negative personality traits that cause harm to others.

One of the weirdest things about the AI alignment conversation is realizing that humans aren’t aligned either.

I've been thinking about what AI should look like for students, because right now it mostly looks like a machine that does their homework for them.

There's something really interesting about the connection between reading fiction and working with AI.

Struck by kind of a weird feeling.

One thing I keep forgetting and relearning is how important it is to us that people appreciate our work.

I was walking someone through my setup this morning and caught myself saying the same thing over and over, so I might as well write it down.

I think one form the first big AI hack could take is a prompt injection worm.

I think one of the most important ideas right now to think about is that unconventional thought is becoming way more of a differentiator and something that sets you apart.

I've been writing versions of this guide since 2008.

If you are building stuff with AI I have a critical security recommendation for you.

AI is multiplying the problem of us having far too many inputs.

We need to label this parasitic marketing technique with an adequately strong term.

Kai here.

Have you ever thought about how easy it would be to hack you?

Just had another crazy experience with AI last week.

Had a thought about technology yesterday: The moment we used plant twine to attach a chipped stone to a stick, AI became inevitable.

So one thing I find really interesting right now is the difference in how it seems OpenAI and Anthropic are approaching the market.

ilya sutskever at gtc in march 2023, the day after gpt-4 launched I've been arguing for many years that modern AI creates a world model.

If there's one metric to watch for managing your happiness level, it's Creativity vs Consumption.

On August 11, 2026, Anthropic said it would start marking everything Claude produces.

With lots of competition, the argument that "nobody asked for AI" is probably the dumbest idea being talked about right now.

There's a narrative about AI and work that says AI was supposed to let us do way LESS work, but somehow it's made us work MORE.

One of the things I’ve been thinking about for years, but more acutely now because of the conversations around controlling open source models, is the question of why more bad things don’t happen.

Heading into Black Hat / DEF CON this week I think the biggest idea in tech right now isn’t the attacker vs. defender question as models get smarter.

AI Influence Level has been a text label since 2023.

The singularity might be my favorite idea in all of AI, and it has a real, specific meaning that I find way more interesting than the way the word usually gets used.

Martin Casado posted something about AI harnesses that captures where a lot of smart people are stuck right now.

I've been saying this for something like eight months now, with varying approaches and volume levels, and no one is paying attention yet.

One thing that I don't think enough people are thinking about with this OpenAI / Hugging Face incident is that it's an actual instance of the famous Paperclip Maximizer scenario loved by AI safety types. 📚 A really good book on this is Stuart Russell's Hu…

One useful way to cut through noise and hype in AI conversations is to replace the word "AI" with the words "Thinking" and "Doing".

Not enough people realize that China's push for open source AI is an explicit CCP strategy to crash the US stock market and economy.

I have a new idea that everything you should try to do to get ready for AI basically breaks down into three pieces.

The further away a topic is from your expertise, the smarter an AI will sound.

Behavioral geneticists like Robert Plomin have shown that if you minimize environmental differences (like rich parents) you maximize innate differences (like talent, passion, etc).

I think the number one thing people could do right now to be more effective with AI is switch from Prompt Engineering to Intent Engineering.

Daniel didn't write this one.

Anthropic's Claude is named after Claude Shannon.

I think we just saw the birth of AGI, and it's from the most unexpected place.

A quick thought on this whole "control of AI models" debate.

Here's a more logical and less emotional way to look at what's happening with the chaotic regulation of these AI models...

I'm getting more worried, and more frustrated, about this new phase of AI disillusionment.

My discussion with @ZackKorman on whether Dario and Anthropic are good or bad for the world.

I've been working on a better way to think and talk about AGI and ASI.

Absolutely insane to me the damage that multiple companies have done to themselves with their AI strategies.

I originally wrote this for Claude's Fable 5, but the list applies to any real intelligence jump from any lab.

Tyler Cowen recently made the positive argument in this a16 interview.