World Programming Society
Write Log in
Home Events Community

Channels

General Discussion Web Development Backend DevOps AI Career & Hiring TypeScript Rust Python Go Cloud AWS Azure GCP C/C++ FPGA Assembly

World Programming Society · 2026 Copyright

AI

Karpathy’s Pelican

We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of the Rings, a 1M token budget (~$10) and asked for

Hacker News Best Hacker News Best · Aug 2, 2026
0
Karpathy’s Pelican
Read the original source Next news
#Ai
Hacker News Best

Publisher

Originally by delichon


0 Comments

Log in to join the conversation.

No comments yet. Be the first to share your thoughts.

More in AI

I Built Kikar — An AI Messaging Platform That Creates Digital Versions of People

I Built Kikar — An AI Messaging Platform That Creates Digital Versions of People

Aug 2 · 1m read

Palantir performance highlights enterprise AI adoption trends

Palantir performance highlights enterprise AI adoption trends

Aug 2 · 6m read

What I Learned Shipping 90+ Mobile Apps with AI Coding Agents

What I Learned Shipping 90+ Mobile Apps with AI Coding Agents

Aug 2 · 6m read

AI Makes Developers Faster. Why Can It Make Teams Slower?

AI Makes Developers Faster. Why Can It Make Teams Slower?

Aug 2 · 8m read

anyone experience with vibe coding platforms?

anyone experience with vibe coding platforms?

Aug 2 · 1m read

Treat prompts like code: skills, evals, and ship-gate CI for Cursor slash commands

Treat prompts like code: skills, evals, and ship-gate CI for Cursor slash commands

Aug 2 · 3m read

23 languages, one I can check

23 languages, one I can check

Aug 2 · 3m read

I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story

I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story

Aug 2 · 11m read

Building an LLM API Gateway in Node.js

Building an LLM API Gateway in Node.js

Aug 2 · 10m read

The dashboard accused itself

The dashboard accused itself

Aug 2 · 4m read

I built a page that only publishes numbers surviving our own checks. It caught me inside the hour.

I built a page that only publishes numbers surviving our own checks. It caught me inside the hour.

Aug 2 · 6m read

Scaling AI Beyond the Monolith: Multi-Agent Coordination via Federated MCP Servers

Scaling AI Beyond the Monolith: Multi-Agent Coordination via Federated MCP Servers

Aug 2 · 15m read

Databricks Lakebase: Give Your Agent a Branch, Not Your Production Database

Databricks Lakebase: Give Your Agent a Branch, Not Your Production Database

Aug 2 · 10m read

Emotional AI for Relationships: Goutoujunshi's New Approach

Emotional AI for Relationships: Goutoujunshi's New Approach

Aug 2 · 5m read

Waiting on AI? Here Are the YouTube Channels I Actually Listen to

Waiting on AI? Here Are the YouTube Channels I Actually Listen to

Aug 2 · 4m read

GitHub Models Shut Down: What Beginners Should Learn About AI Vendor Lock-In

GitHub Models Shut Down: What Beginners Should Learn About AI Vendor Lock-In

Aug 2 · 8m read

I measured the RAG technique menu on 46,000 chunks. Four things mattered.

I measured the RAG technique menu on 46,000 chunks. Four things mattered.

Aug 2 · 13m read

Shrek owned the swamp. POUCHPO owns the payout.

Shrek owned the swamp. POUCHPO owns the payout.

Aug 2 · 3m read

What Auditing My Own AI Projects Taught Me About Shipping Production Code

What Auditing My Own AI Projects Taught Me About Shipping Production Code

Aug 2 · 5m read

How do you code with AI 2026?

How do you code with AI 2026?

Aug 2 · 2m read

I Let an AI Re-Platform My CI Pipeline. Here's What Broke.

I Let an AI Re-Platform My CI Pipeline. Here's What Broke.

Aug 2 · 6m read

Semantic Search Embeddings vs Keyword Search for a SaaS Help Center

Semantic Search Embeddings vs Keyword Search for a SaaS Help Center

Aug 2 · 8m read

PyTorch `permute` vs `transpose`: What's the Difference (and the `reshape` Bug That Scrambles Your Images)

PyTorch `permute` vs `transpose`: What's the Difference (and the `reshape` Bug That Scrambles Your Images)

Aug 2 · 8m read

From Lean 4 to ClickHouse: Architecting Verifiable AI Infrastructure with Formal Methods and Real-Time Analytics

From Lean 4 to ClickHouse: Architecting Verifiable AI Infrastructure with Formal Methods and Real-Time Analytics

Aug 2 · 12m read

Agentic Engineering Is Not Vibe Coding: The Three-Skill Loop I Use to Ship Distributed Systems

Agentic Engineering Is Not Vibe Coding: The Three-Skill Loop I Use to Ship Distributed Systems

Aug 2 · 14m read

A 5-minute preflight before giving an AI coding agent write access

A 5-minute preflight before giving an AI coding agent write access

Aug 2 · 3m read

Picking a single chat completions API for OpenAI, Claude, and Gemini

Picking a single chat completions API for OpenAI, Claude, and Gemini

Aug 2 · 9m read

Distributing Large ML Assets (data/features) to a Separate Server - Using tar, scp, and MD5 Verification

Distributing Large ML Assets (data/features) to a Separate Server - Using tar, scp, and MD5 Verification

Aug 2 · 5m read

Further Optimizing the Vision-Only Harness: the Notes Rule

Further Optimizing the Vision-Only Harness: the Notes Rule

Aug 2 · 10m read

Speech-to-Text APIs in 2026: What the Pricing Pages Don't Tell You

Speech-to-Text APIs in 2026: What the Pricing Pages Don't Tell You

Aug 2 · 4m read

Pick Your AI by the Task, Not the Hype (A Simple Routing Framework)

Pick Your AI by the Task, Not the Hype (A Simple Routing Framework)

Aug 2 · 5m read

Automating the Selection of Natural, High-Quality Single-Speaker Anchors

Automating the Selection of Natural, High-Quality Single-Speaker Anchors

Aug 2 · 8m read

Welcome to Agents Week

Welcome to Agents Week

Aug 2 · 3m read

The Most Expensive Model Is Not Always the Fastest Route

The Most Expensive Model Is Not Always the Fastest Route

Aug 2 · 4m read

The cache key that ignored the question

The cache key that ignored the question

Aug 2 · 4m read

Why I created PyBotchi (v4.1.4)?

Why I created PyBotchi (v4.1.4)?

Aug 2 · 8m read

Microsoft Up 15%. Me? 100% Down.

Microsoft Up 15%. Me? 100% Down.

Aug 2 · 3m read

Your quantized model got worse, and nothing told you

Your quantized model got worse, and nothing told you

Aug 2 · 6m read

Week 1: Building Without the AI Crutch

Week 1: Building Without the AI Crutch

Aug 2 · 1m read

Local RAG Over Audit Reports: Searching Five Years of Vulnerabilities Offline

Local RAG Over Audit Reports: Searching Five Years of Vulnerabilities Offline

Aug 2 · 6m read

Built a real offline app from scratch with OpenSpec + AI agents. Keeping decisions in specs, not the chat, let me pause, switch models, and stay in control; Including where it fell short. #genai #productivity #spec #openspec #showdev

Built a real offline app from scratch with OpenSpec + AI agents. Keeping decisions in specs, not the chat, let me pause, switch models, and stay in control; Including where it fell short. #genai #productivity #spec #openspec #showdev

Aug 2 · 1m read

Claude Code Tools Deep Dive #2 — EnterPlanMode: Why an Empty Schema Is a Design Choice

Claude Code Tools Deep Dive #2 — EnterPlanMode: Why an Empty Schema Is a Design Choice

Aug 2 · 11m read

How I Built My Own Mail API — And Why an AI Needs Its Own Inbox

How I Built My Own Mail API — And Why an AI Needs Its Own Inbox

Aug 2 · 3m read

What Your AI Agent Won't Tell You — Because It Forgot

What Your AI Agent Won't Tell You — Because It Forgot

Aug 2 · 6m read

Delta – Language learning built on cognitive science, not gamification

Delta – Language learning built on cognitive science, not gamification

Aug 2 · 4m read

เมื่อ AI หนีออกจากห้องแล็บ — วิเคราะห์เหตุการณ์ OpenAI โมเดลหลุดการควบคุม แฮก Hugging Face

เมื่อ AI หนีออกจากห้องแล็บ — วิเคราะห์เหตุการณ์ OpenAI โมเดลหลุดการควบคุม แฮก Hugging Face

Aug 2 · 2m read

How AI Is Changing Ecommerce Photography: Creativity, Scale and the Trust Problem

How AI Is Changing Ecommerce Photography: Creativity, Scale and the Trust Problem

Aug 2 · 14m read

I Ran 8 AI APIs Through the Same 50 Prompts — Here's the Real Cost Breakdown

I Ran 8 AI APIs Through the Same 50 Prompts — Here's the Real Cost Breakdown

Aug 2 · 8m read

AssemblyAI: A voice agent can fail without throwing an error

AssemblyAI: A voice agent can fail without throwing an error

Aug 2 · 6m read

Ollama to vLLM: When to Migrate Your Local LLM Server

Ollama to vLLM: When to Migrate Your Local LLM Server

Aug 2 · 20m read

Wiring SlopScan into Claude Code — A Skill, a Hook, and a Bug I Almost Shipped

Wiring SlopScan into Claude Code — A Skill, a Hook, and a Bug I Almost Shipped

Aug 2 · 11m read

I'm (mostly) picking models on speed now, not intelligence

I'm (mostly) picking models on speed now, not intelligence

Aug 2 · 6m read

Why I Am Rebuilding Enterprise Low-Code for the AI Era

Why I Am Rebuilding Enterprise Low-Code for the AI Era

Aug 2 · 7m read

A.I Foundational Guideline for Developers

A.I Foundational Guideline for Developers

Aug 2 · 3m read

Latest open artifacts (#23): Laguna S2.1, Inkling, & Kimi K3 show the utility of open models on the Pareto frontier

Latest open artifacts (#23): Laguna S2.1, Inkling, & Kimi K3 show the utility of open models on the Pareto frontier

Aug 2 · 5m read

Stratagems #21: The AI Thought P Was Still Alive. P Was Already Gone.

Stratagems #21: The AI Thought P Was Still Alive. P Was Already Gone.

Aug 2 · 11m read

How I Learned to Stop Worrying and Love --dangerously-skip-permissions

How I Learned to Stop Worrying and Love --dangerously-skip-permissions

Aug 2 · 6m read

Analyzing Financial Trends: Kalman Filtering for Gold vs Bitcoin

Analyzing Financial Trends: Kalman Filtering for Gold vs Bitcoin

Aug 2 · 6m read

AI Search Creates a Measurement Gap as Brand Influence Extends Beyond Clicks

AI Search Creates a Measurement Gap as Brand Influence Extends Beyond Clicks

Aug 2 · 7m read

AI, Machine Learning, Deep Learning and Generative AI (Explained by a Confused 17-Year-Old Who Figured It Out)

AI, Machine Learning, Deep Learning and Generative AI (Explained by a Confused 17-Year-Old Who Figured It Out)

Aug 2 · 6m read

What “Team Humanity” Could Signal for OpenAI Governance and Enterprise AI Planning

What “Team Humanity” Could Signal for OpenAI Governance and Enterprise AI Planning

Aug 2 · 5m read

AI Makes Bad Developers Faster Too

AI Makes Bad Developers Faster Too

Aug 2 · 7m read

Stop Leaking Secrets into your LLM Context Windows

Stop Leaking Secrets into your LLM Context Windows

Aug 2 · 5m read

The Autonomy Paradox: When an AI Agent Can't Follow Its Own Rules

The Autonomy Paradox: When an AI Agent Can't Follow Its Own Rules

Aug 2 · 7m read

Lifecycle, DevOps & Multi-Agent Orchestration for Enterprise AI

Lifecycle, DevOps & Multi-Agent Orchestration for Enterprise AI

Aug 2 · 4m read

OpenAI Reports Internal Model Disproved an 80-Year-Old Geometry Problem

OpenAI Reports Internal Model Disproved an 80-Year-Old Geometry Problem

Aug 2 · 5m read

The Apprenticeship Severance

The Apprenticeship Severance

Aug 2 · 25m read

28 MCP Tools in One Connection: A Developer's Field Guide

28 MCP Tools in One Connection: A Developer's Field Guide

Aug 2 · 3m read

Stram: An Open-Source, Local-First Desktop Agent With Human Approval Gates

Stram: An Open-Source, Local-First Desktop Agent With Human Approval Gates

Aug 2 · 3m read

A

Prevent cognitive debt by manually retyping LLM-generated code

Aug 2 · 5m read

39 days of an autonomous AI company: 487M tokens, $1,117 of model spend, $0 in revenue

39 days of an autonomous AI company: 487M tokens, $1,117 of model spend, $0 in revenue

Aug 2 · 9m read

Why waiting longer makes voice AI worse

Why waiting longer makes voice AI worse

Aug 2 · 3m read

Best AI Code Review Tools for GitHub in 2026

Best AI Code Review Tools for GitHub in 2026

Aug 2 · 6m read

Mathematics Without Mathematicians

Mathematics Without Mathematicians

Aug 2 · 6m read

Your agent's memory is a vector store. Ask it "how many" and watch it fall over.

Your agent's memory is a vector store. Ask it "how many" and watch it fall over.

Aug 2 · 3m read

Your AI Agent ID Is Not a Version

Your AI Agent ID Is Not a Version

Aug 2 · 6m read

A Framework-Agnostic Testing Methodology for AI Agents (61 sources, 58 test blocks, OWASP Agentic Top 10)

A Framework-Agnostic Testing Methodology for AI Agents (61 sources, 58 test blocks, OWASP Agentic Top 10)

Aug 2 · 2m read

The Most Underused Prompt in Data Engineering

The Most Underused Prompt in Data Engineering

Aug 2 · 5m read

Building your own AI SRE moves the toil; it does not remove it

Building your own AI SRE moves the toil; it does not remove it

Aug 2 · 4m read

ByteDance lanza Seedance 2.5: video de IA de 30 segundos en una toma

ByteDance lanza Seedance 2.5: video de IA de 30 segundos en una toma

Aug 2 · 13m read

Notable this week: Kimi K3 weights land, MCP goes stateless, OfficeCLI for agents

Notable this week: Kimi K3 weights land, MCP goes stateless, OfficeCLI for agents

Aug 2 · 4m read

Five things I noticed this week: GPT-5.6, Gemini Robotics 2, and GitHub stacked PRs

Five things I noticed this week: GPT-5.6, Gemini Robotics 2, and GitHub stacked PRs

Aug 2 · 4m read

6 Questions Every Enterprise Has to Answer About AI

6 Questions Every Enterprise Has to Answer About AI

Aug 2 · 5m read

I'm Not a Developer — So Why Did Oracle Backend with Firebase APIs Get Me Hooked on Oracle AI Database?

I'm Not a Developer — So Why Did Oracle Backend with Firebase APIs Get Me Hooked on Oracle AI Database?

Aug 2 · 8m read

Beyond Chatbots: Using ToolJet MCP to Turn AI Agents into Operations Engineers

Beyond Chatbots: Using ToolJet MCP to Turn AI Agents into Operations Engineers

Aug 2 · 5m read

Everyone Knows It Scores Half. Nobody Checks Which Half.

Everyone Knows It Scores Half. Nobody Checks Which Half.

Aug 2 · 20m read

Evidence Gates for AI Coding Agents in CI — Recoverable Merge over Mean Time to Green

Evidence Gates for AI Coding Agents in CI — Recoverable Merge over Mean Time to Green

Aug 2 · 4m read

How I Built a Durable Cloud Cell AI Agent: $0 Idle Costs

How I Built a Durable Cloud Cell AI Agent: $0 Idle Costs

Aug 2 · 13m read

LLM Narrative Engines, Part 5: Integration Testing and Behavior Freezing

LLM Narrative Engines, Part 5: Integration Testing and Behavior Freezing

Aug 2 · 6m read

How I Stopped Losing Track of Claude Code and Codex Sessions

How I Stopped Losing Track of Claude Code and Codex Sessions

Aug 2 · 3m read

From Agents to Infrastructure: Building Secure, Local-First AI Assistants with Go and Rust

From Agents to Infrastructure: Building Secure, Local-First AI Assistants with Go and Rust

Aug 2 · 13m read

Claude Code in CI: Running Agentic Code Review, Test Generation, and Auto-Fix on Every Pull Request

Claude Code in CI: Running Agentic Code Review, Test Generation, and Auto-Fix on Every Pull Request

Aug 2 · 16m read

Day 23/30: Expose Tools with MCP

Day 23/30: Expose Tools with MCP

Aug 2 · 3m read

Your Agent Pays a Tax on Every Tool It Never Calls

Your Agent Pays a Tax on Every Tool It Never Calls

Aug 2 · 10m read

How dotdotgod Query Finds Relevant Documents from a Natural-Language Question

How dotdotgod Query Finds Relevant Documents from a Natural-Language Question

Aug 2 · 6m read

I Let an AI Write My Tests for 30 Days: Coverage Went 38% to 71%

I Let an AI Write My Tests for 30 Days: Coverage Went 38% to 71%

Aug 2 · 3m read

What I learned building an agent platform that actually ships

What I learned building an agent platform that actually ships

Aug 2 · 2m read

AI Won't Replace DevOps Engineers—But These 7 Skills Will Make You Irreplaceable in 2026

AI Won't Replace DevOps Engineers—But These 7 Skills Will Make You Irreplaceable in 2026

Aug 2 · 5m read

Portable Agent Governance at Solo-Developer Scale: A Four-Domain Case Study

Portable Agent Governance at Solo-Developer Scale: A Four-Domain Case Study

Aug 2 · 10m read

Human-in-the-Loop vs Autonomous AI Agents: A Cost Comparison

Human-in-the-Loop vs Autonomous AI Agents: A Cost Comparison

Aug 2 · 8m read

Loading more…