AI-Toolbox: Who’s building it? Andrej Karpathy recently tweeted this: “Input optional product Don’t ask your users for... 2024-10-3 08:0:0 | 阅读: 1 | 收藏 | Joseph Thacker - josephthacker.com prompts profession toolbox barrier creativity

The Data Wall, Agents, and Planning-Based Evals I’ve been thinking a lot about the whole “data wall” thing with LLMs lately. It’s the idea tha... 2024-8-22 08:0:0 | 阅读: 0 | 收藏 | Joseph Thacker - josephthacker.com planning llm answers wall agentic

Internal Monologue Capture I can’t stop thinking about a new concept AI applications could benefit from. I’m calling it i... 2024-8-1 08:0:0 | 阅读: 7 | 收藏 | Joseph Thacker - josephthacker.com monologue thoughts imagine expertise constantly

Unleashing Claude 3.5 Sonnet As A Hacker Claude 3.5 was recently released, and it’s a clear step up from any other model currently avai... 2024-6-29 08:0:0 | 阅读: 13 | 收藏 | Joseph Thacker - josephthacker.com claude sonnet jailbreak godmode payload

Defining Real AI Risks Yann LeCun is making the same mistake Marc Andreesen makes when considering AI risk. They don’... 2024-5-19 08:0:0 | 阅读: 7 | 收藏 | Joseph Thacker - josephthacker.com alignment injection security bias llms

Empowering Long-Running AI Agents with Timers There’s been a lot of discussion lately about how AI struggles with long-running tasks. And it... 2024-5-16 08:0:0 | 阅读: 3 | 收藏 | Joseph Thacker - josephthacker.com wake constantly snooze posting narrow

GPT-4o: Actually Good Multimodal AI OpenAI just made a big move in the AI space with the release of GPT-4o (“o” stands for “omni”)... 2024-5-14 08:0:0 | 阅读: 1 | 收藏 | Joseph Thacker - josephthacker.com gpt 4o copilot adapt claude

The Meta AI Ray-Bans Are Awesome My wife got me the new Meta AI Ray-Bans as an early birthday present, and I have to say, they... 2024-5-8 08:0:0 | 阅读: 3 | 收藏 | Joseph Thacker - josephthacker.com llm ray hey bans sunglasses

The Three Categories of AI Agent Auth As I’ve been discussing AI agent authentication with some brilliant people in San Fran this we... 2024-5-8 08:0:0 | 阅读: 4 | 收藏 | Joseph Thacker - josephthacker.com agents b2c b2b workflows phones

assumptions_made Pitching the assumptions_made metadata field for AI AgentsI’ve been thinking a lot lately ab... 2024-5-4 08:0:0 | 阅读: 4 | 收藏 | Joseph Thacker - josephthacker.com assumptions llm agents dinner llms

Rabbit r1: Innovative Device, Security Concerns Hey friends, let me start by saying - I’m actually really excited about Rabbit’s new r1 device... 2024-4-26 08:0:0 | 阅读: 1 | 收藏 | Joseph Thacker - josephthacker.com vnc rabbit security agents facebook

Incremental Learning LLM Pattern I had a cool idea for a new LLM-based application this week that I think has a ton of potentia... 2024-4-24 08:0:0 | 阅读: 2 | 收藏 | Joseph Thacker - josephthacker.com wikis progression imagine incremental feeding

Positive Mental Framing Examples Mental framing can be powerful. Framing is the way you percieve a situation and the story you... 2024-3-5 08:0:0 | 阅读: 10 | 收藏 | Joseph Thacker - josephthacker.com reframing framing mental cold annoyed

All About Hackbots: AI Agents That Hack A hackbot is the term I’ve settled on using for AI systems with hacking abilities. I saw the t... 2024-2-21 08:0:0 | 阅读: 12 | 收藏 | Joseph Thacker - josephthacker.com hackbot hackbots hero dreams

From Concept to Capability: Required Security Changes for Secure AI Agents Capable AI agents will require new infrastructure. In order to expedite the utility of digital... 2024-2-5 08:0:0 | 阅读: 8 | 收藏 | Joseph Thacker - josephthacker.com agents menus ification restaurant

Adapting to Advancements I recently read Amusing Ourselves to death. It was a fantastic book; way ahead of its time. I... 2023-11-29 08:0:0 | 阅读: 9 | 收藏 | Joseph Thacker - josephthacker.com ourselves healthy amusing sharp

AI Hacking Agents Will Outperform Humans I believe AI Agents will outhack humans. Here’s why.The Logic for HowThere’s a finite numbe... 2023-11-8 08:0:0 | 阅读: 12 | 收藏 | Joseph Thacker - josephthacker.com agents humans surpass scaled

Beyond the Blog: More AI and Hacking Content If you read my posts on here, you enjoy the same things I do. So, I wanted to let you know abo... 2023-11-4 08:0:0 | 阅读: 9 | 收藏 | Joseph Thacker - josephthacker.com security ethical injection reward hackerone

AI Security Has Serious Terminology Issues AI Security is an emerging field. I LOVE being at the forefront of something so transformative... 2023-10-16 08:0:0 | 阅读: 12 | 收藏 | Joseph Thacker - josephthacker.com security clarity talking unclear llm

Jailbreaking Humans vs Jailbreaking LLMs “Jailbreaking” an LLM and convincing it to tell you things it’s not supposed to is very simila... 2023-10-11 08:0:0 | 阅读: 10 | 收藏 | Joseph Thacker - josephthacker.com llms windows humans infinite