WeatherNext 2 — Google DeepMind — DeepMind’s WeatherNext 2 is a useful look at how AI forecasting is improving, especially on accuracy. Worth skimming if you care about what’s changing in short-term weather prediction and its practical limits. https://deepmind.google/science/weathernext/

Tomorrow — bring your questions; we'll keep it hands-on. LangChain for Generative AI Pipelines. Register: learning.oreilly.com/live-events/-/0642572002...

Tomorrow — bring your questions; we'll keep it hands-on. LangChain for Generative AI Pipelines. Register: learning.oreilly.com/live-events/-/0642572002...

Scientists stunned by ‘fundamentally new way’ life produces DNA — Worth a read; the useful bits are usually in the examples and e… www.science.org/content/article/scientis...

Stanford's AI Index for 2026 Shows the State of AI - IEEE Spectrum — Stanford’s 2026 AI Index is a useful snapshot of where AI is heading, with clear numbers on compute growth, emissions, and shifting public trust. Worth skimming if you want data to… https://spectrum.ieee.org/state-of-ai-index-2026

What are skiplists good for? | Antithesis Blog — A clear, practical look at where skiplists actually shine, plus a grounded comparison to what you’d otherwise end up doing with lots of SQL JOINs. Worth a read if you’re weighing data structure choices for… https://antithesis.com/blog/2026/skiptrees/

Building Large Language Models (LLMs) A clear, practical overview of what goes into building an LLM, from data and training to evaluation and deployment. Worth a skim if you want the end-to-end picture without getting lost in theory.

Building Large Language Models (LLMs) — A clear, practical overview of what goes into building an LLM, from data and training to evaluation and deployment. Worth a skim if you want the end-to-end picture without getting lost in theory. https://m.youtube.com/watch?v=9vM4p9NN0Ts

How We Build Effective Agents: Barry Zhang, Anthropic A clear look at the practical engineering choices behind building effective AI agents, straight from Anthropic. Worth a skim if you’re thinking about tool use, evaluation, and where agent reliability actually comes from.

How We Build Effective Agents: Barry Zhang, Anthropic — A clear look at the practical engineering choices behind building effective AI agents, straight from Anthropic. Worth a skim if you’re thinking about tool use, evaluation, and where agent reliability… https://m.youtube.com/watch?v=D7_ipDqhtwk

Maximum entropy temporal networks — Worth a read; the useful bits are usually in the examples and edge cases. https://journals.aps.org/pre/abstract/10.1103/78vv-hs72

[2604.14228] Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems — A useful overview of how Claude Code frames the design space for agentic systems—what to build into the agent versus the environment. Worth a skim if you’re thinking about… https://arxiv.org/abs/2604.14228

[2604.13018] Toward Autonomous Long-Horizon Engineering for ML Research — A useful look at what it would take for agents to handle long-horizon engineering work in ML research, beyond short coding tasks. Worth skimming for the problem framing and where current… https://arxiv.org/abs/2604.13018

The Abstraction Fallacy: Why AI Can Simulate But Not Instantiate Consciousness — Worth a read; the useful bits are usually in the examples and edge cases. https://philpapers.org/rec/LERTAF

I Measured Claude 4.7's New Tokenizer. Here's What It Costs You. — Useful reality check on Claude 4.7’s tokenizer: the author measured ~1.47× token inflation on real text versus the 1.0–1.35× range in th… www.claudecodecamp.com/p/i-measured-claude-4-7-...

Introduction to Spherical Harmonics for Graphics Programmers — Clear, programmer-focused intro to spherical harmonics, with just enough math to make the common graphics uses (especially lighting approximation) feel approachable. Good primer before diving into… https://gpfault.net/posts/sph.html

Slop is text you haven't read, not text you haven't written — Good framing on the “slop” debate: the real failure mode is sharing text you didn’t read, regardless of whether a human or an LLM wrote it. A useful reminder to treat review… dwyer.co.za/static/slop-is-text-you-...

David J. Chalmers, What we talk to when we talk to language models - PhilArchive — Chalmers offers a clear way to think about what we’re actually engaging with when we “talk” to an LLM, especially around whether it makes sense to attribute mental states. Useful… https://philarchive.org/rec/CHAWWT-8

Tool calling, open source, and the M×N problem — Clear overview of why tool calling is easy with closed models but messy in open-source setups—the M×N integration problem is real. Worth a read if you’re building function/tool… www.thetypicalset.com/blog/grammar-parser-main...

One week out — last chance to plan it into your calendar. LangChain for Generative AI Pipelines. Register: learning.oreilly.com/live-events/-/0642572002...

One week out — last chance to plan it into your calendar. LangChain for Generative AI Pipelines. Register: learning.oreilly.com/live-events/-/0642572002...

Two weeks out — here's the practical angle (tools, patterns, gotchas). Claude API for Python Developers. Register: learning.oreilly.com/live-events/-/0642572255...

Two weeks out — here's the practical angle (tools, patterns, gotchas). Claude API for Python Developers. Register: learning.oreilly.com/live-events/-/0642572255...

My understanding is that they’re essentially cron jobs on steroids

Automate work with routines - Claude Code Docs — Clear overview of Claude Code routines—how to schedule jobs, trigger them via API, or hook into GitHub events. Practical starting point if you’re looking to automate repeatable workflows with managed… https://code.claude.com/docs/en/routines

Stanford Artificial Intelligence Index Report 2026 — Useful snapshot of where AI is actually heading in 2026—metrics on research, investment, regulation, and real-world adoption in one place. Worth skimming for grounded numbers you can… hai.stanford.edu/assets/files/ai_index_re...

[2604.07709] IatroBench: Pre-Registered Evidence of Iatrogenic Harm from AI Safety Measures — Worth a look if you care about the trade-offs in AI safety: this paper sets up a pre-registered benchmark to test whether safety interventions can inadvertently cause… https://arxiv.org/abs/2604.07709

[2604.08224] Externalization in LLM Agents: A Unified Review of Memory, Skills, Protocols and Harness Engineering — Useful unified review of how agent “externalization” actually gets built in practice—memory stores, skill libraries, protocols, and harness tooling.… https://arxiv.org/abs/2604.08224

Center for Responsible, Decentralized Intelligence at Berkeley — A useful reality check on AI agent benchmarks: Berkeley researchers show how top leaderboards can be gamed to get near-perfect scores without doing the task. Worth reading if… rdi.berkeley.edu/blog/trustworthy-benchma...

[2604.06425] Neural Computers — Useful overview of the “neural computers” idea—how models can be structured to compute with learned representations rather than just pattern-match. Worth a skim if you’re tracking where architecture and algorithm design are converging. https://arxiv.org/abs/2604.06425

Anthropic Will Use CoreWeave’s AI Capacity to Power Claude - Bloomberg — Worth a read; the useful bits are usually in the examples and edge cases. www.bloomberg.com/news/articles/2026-04-10...

Towards transparency and knowledge exchange in AI-assisted data analysis code generation | Nature Computational Science — A concise perspective on why AI-generated analysis code needs clearer provenance, documentation, and sharing norms to be… https://www.nature.com/articles/s43588-025-00781-1