danielmiessler.bsky.social
AI / Security Researcher and Entrepreneur.
Founder/CEO of Unsupervised Learning.
Building AI that upgrades humans.
260 posts
2,545 followers
88 following
Regular Contributor
Active Commenter
comment in response to
post
So glad you’re finding it useful!
comment in response to
post
It’s not easy. You have to try to infer what they’re trying to build from their actions.
And I differentiate bad actions from a bad ideal world. Someone can be an asshole and still be trying to build a world most people would enjoy living in.
The trick is telling the difference.
comment in response to
post
Same.
And I sometimes go to people I know supported the current president, and just check in...
Is this what you voted for?
Too often they say yes, though, so I do that less now.
: (
comment in response to
post
That's why an AGI definition based on human-work-replacement makes the most sense.
If you see a way for me to improve the definition, or you know of a better one, let's discuss.
comment in response to
post
I think in the next few years this will become the most used type of definition, i.e., a real-world capability relative to humans.
AI doesn't matter because of technology itself. Who cares about random technical benchmarks? AI matters because of PEOPLE, and how it affects them.
comment in response to
post
Looking at our criteria:
1. We'll know when it happens—because AI will be used to actually replace the work done by an average knowledge worker. And…
2. We know why it matters—because there are (according to Forbes) around one BILLION knowledge workers on the planet.
comment in response to
post
So instead of arguing about the definition of General Intelligence, how about we just take something we KNOW has it—which is an actual human doing knowledge work professionally—and use that as the standard?
comment in response to
post
-> The ability for an AI—whether a model, a product, or a system—to perform the work of an average US-based knowledge worker from 2022.
That's it. It can replace a decent knowledge worker from before modern AI.
comment in response to
post
#FF
@danielmiessler.bsky.social
@anmolagarwal.bsky.social
@angietechcafe.bsky.social
@leonvqz.com
@jamellebouie.net
@bsky.app
@3ncr1pt3d.bsky.social
@miketalonnyc.bsky.social
@ellearmageddon.bsky.social
@hacks4pancakes.com
@julien.bcksec.com
@malwarejake.bsky.social
@alyssam-infosec.com
comment in response to
post
🙏For more breakdowns, news, and analysis, you can follow me @danielmiessler.bsky.social to never miss an update.
➡️And if you want to get even more analysis of what’s happening in security, AI, and tech, you should follow the conversation at:
danielmiessler.com
comment in response to
post
📢 Sponsored by @WorkOS
🛡️ Protect Your App with WorkOS Radar: Block fake signups, bots, and brute force attacks with advanced device fingerprinting. Keep your app secure & your users safe with a simple API.
workos.com/radar?utm_me...
comment in response to
post
📢 Sponsored by Flashpoint
🚨 Ransomware Survival Guide 2025: 4,500+ attacks hit in 2024—53% targeting U.S. companies. Learn how ransomware ops use phishing, infostealers & exploits, and how to defend against them.
📘 Get the free report:
go.flashpoint.io/ransomware-s...
comment in response to
post
🔍Discoveries:
🔥 UV as a Python Shebang @akrabat.com
🕸️ AI Crawlers vs. Tarpits
🔍 Deep Research in ChatGPT
🎭 Death of Subculture: How commodification is erasing authenticity. @justinmcguirk.bsky.social
📺 YouTube Downloader: High-quality, ad-free downloads.
comment in response to
post
Thank you, Philip!
comment in response to
post
@jacenrkohler.net
@malwaretech.com
@tracketpacer.bsky.social
@racheltobac.bsky.social
@s0ciabl3ng.bsky.social
@sushicomabacate.com
@scalzi.com
@noid.hotunixsluts.org
@michaeljohan.bsky.social
@danielmiessler.bsky.social
@marczammit.bsky.social
@insider.phd
@angietechcafe.bsky.social
comment in response to
post
Brilliant! Espanso? Raycast?
comment in response to
post
📢 Sponsored by @dropzoneai.bsky.social
⏱️ The SOC Speed Metric You Need: MTTC reveals true SOC efficiency. 90% of SOCs drown in alerts—Dropzone AI reduces MTTC with AI-driven automation.
📘 Get the free eBook: content.dropzone.ai/why-speed-in...
comment in response to
post
📢 Sponsored by @paloaltonetworks.bsky.social
🎙️ Listen to Threat Vector – the official Palo Alto Networks podcast on cybersecurity threats, protection strategies, and industry trends. Featuring expert insights to help safeguard your org.
Tune in: www.paloaltonetworks.com/podcasts/thr...
comment in response to
post
🔍 Discoveries
- 🔥 Cline: The best AI assistant I’ve used so far.
- 🤖 Try Deepseek: Watch it think in real-time via Ollama. @ollama.bsky.social
- 🌐 LangChain: New local web research assistant. @langchain.bsky.social
comment in response to
post
Job loss is one, but so is a massive increase in security vulnerabilities.
#AI #Cybersecurity #AIMisconfigurations #ApplicationSecurity #APISecurity #TechRisks #AIApplications #DigitalVulnerabilities #FutureOfTech #TechLeadership
comment in response to
post
But they will not be asked to do so, because the only goal of the person barking commands at the AI will be to make it happen and to ship it.
I am very pro AI and I can’t wait to see all the things that we make with it, but there will be extra realities.
comment in response to
post
They don’t understand the underlying Webb technologies, the underlying API, the underlying programming languages, or how to secure any of those things .
And the AI might actually be halfway decent at securing those things, if they were asked to do so.
comment in response to
post
It is super exciting that millions of non-developers are about to be able to make things, but there is a lot of danger that comes with that because they have no idea what they are actually making.
comment in response to
post
These things are basically Jenga towers of misconfiguration and sloppiness, and it will cause a supernova of vulnerabilities—especially in web applications and APIs.
comment in response to
post
📢Sponsored by @IntezerLabs
🤖 Extend Your SOC with AI:
-Eliminate alert fatigue with accurate, autonomous triage
-Investigate alerts in 2 minutes on average
-Catch hidden threats with AI that escalates what matters
Transform your SOC with Intezer: intezer.com?utm_campaign...
comment in response to
post
📢Sponsored by @paloaltonetworks.bsky.social
🌩️ Symphony 2025: See the future of real-time cloud security.
First look at bold cloud security innovations
Exclusive Unit 42® threat insights
Game-changing Cortex® demos
Actionable takeaways for your SOC
start.paloaltonetworks.com/symphony2025...
comment in response to
post
🔍 Discoveries
⭐ Favorite Stuff 2024 @JosephThacker
📝 Use Raycast to avoid name mishaps @sulco.dev
💼 Before Quitting: Lessons learned before leaving a job. @michaeldrogalis.bsky.social
💡 Hypothesis Sheets: A framework for startup ideas. @michaelrbock.com