RSSAmplifier

Blog

Adam Jones's Blog

Adam Jones's Blog

adamjones.me ↗RSS feed ↗119 posts

Latest posts

Thinking tool: Defend slack

Thinking tool: Ask if the answer changes the decision

Thinking tool: Be Pareto-best in the world

For better outcomes: we can slow AI capabilities AND accelerate AI safety

RL environments I'd find interesting

Thinking tool: Find problems, not applications for solutions

Thinking tool: Pick up the trillion-dollar bill

Thinking tool: Hunt for 100x opportunities

Thinking tool: Accept redundancy

Thinking tool: Musk's 5 step design process (question, delete, simplify, accelerate, automate)

Thinking tool: Framestorming

Thinking tool: Optimise only the constraint by finding your bottleneck

Thinking tool: Create surface area for luck

Thinking tool: Backwards chaining

Effective Patterns for Advanced MCP Usage

Suddenly becoming vegetarian, and maybe even vegan???

Run better job interviews by viewing them as measurement tools

Interview questions that find high agency people

Do structured interviews actually work?

CLAUDE.md Considered Harmful

Introducing: 'highvol'-tagged blog posts

What working at Palantir was actually like

Announcing... AdamCon 2026!

A guide to building useful AI tools for AI safety

Good MCP servers are designed for composition

I'm hiring: Software Engineer, RL Data at Anthropic

How we contain Claude across products

Running LLMs Locally in 2026: Local models got ~2.8x faster in a year

What's the best dictation app? I tried 21 Wispr Flow alternatives to find out.

Working at Anthropic FAQs

How a coalition of democracies solves AGI's biggest risks

UBI doesn't solve AI power concentration

Open source doesn't solve AI power concentration

How AGI could kill US democracy

Give AI companies something to aim for: The case for beneficial capability evals

Langfuse Context: All things MCP with Adam Jones (Tech Lead at Anthropic)

MCP Connect with Anthropic, Zoopla & Alpic | London edition

Code Execution with MCP: Fix Tool Token Bloat (Adam Jones, Anthropic)

Code execution with MCP: Building more efficient agents

The Model Context Protocol: Connecting AI to Everything - Adam Jones, Anthropic

An AI safety plan that might work: an international coalition governing AGI

What should the UK be doing to make AGI go well?

Announcing... AdamCon 2025!

Overcoming problems with compute thresholds for AI regulation

AI Model Thresholds: How Governments Can Identify Frontier AI Systems

5 ways to develop an AI safety plan or field strategy

Quantitative models of AI-driven bioterrorism and lab leak biorisk

What makes a good AI safety plan?

Key paths, plans and strategies to AI safety success

How you can get more value from me reviewing your work