- 1 hour 22 minutes189: Agentic Loops189: Agentic Loops
Intro topic: Building in plain language
News/Links:- RAM prices going through the roof . Tom’s Hardware
- Mario meet Pareto
- Qwen 3.8 versus Muse Glimmer
- Everyone should know SIMD
- Patrick
- Waves in an Impossible Sea by Matt Strassler
- Jason
- Diffusion World Modeling
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show- Patrick
- Nubby’s Number Factory
- Jason
- Opencode https://opencode.ai/
- Opencode https://opencode.ai/
- Managing llm uncertainty
- Polling / Timed Loops
- Intelligent cronjobs
- Autoresearch
- Andrej Karpathy
- Recursive Self Improvement
- Verifiers
- Terminating conditions
- Managing context
- Subagents
- Best Practices
- Watchdogs to correct loops that are stuck/aborted
24 August 2026, 4:38 pm - 1 hour 36 minutes188: World ModelsIntro topic: Running
News/Links:- Flow matching versus diffusion models
- OpenCV 5
- Claude fable beats pokemon with no harness
- Can the stockmarket swallow Anthropic, SpaceX and OpenAI?
- Patrick
- Strength of the Few - James Islington
- Jason
- Descender - Jeff Lemire
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show- Patrick
- No Man’s Sky
- Jason
- Paperlib https://paperlib.app/en/
- Paperlib https://paperlib.app/en/
- Making decisions with AI
- Action-Value (called a Q model): What is the long-term value of making a decision at a position
- Policy: What action should I take (must be a distribution)
- Value (called a V model): What is the value of a position (depends on policy)
- Advantage/Disadvantage: difference in value given two policies
- When advantage is +, do that more.
- Model-Free
- Look at the current situation and suggest an action
- Run that action in the real world and measure the effect
- Use that measurement to suggest better actions next time
- Model-Based
- Observe rollouts (sequences of situations) and learn the dynamics
- Choose an action, use your dynamics model to measure the consequence
- Potentially do MPC (try many actions and choose the best)
- World Models
- Observe many many rollouts and learn a full forward model (how to create the input in the future)
- Train a policy & value inside the world model
- Deploy the policy and fine-tune based on the real world
9 July 2026, 4:22 pm - 1 hour 38 minutes187: Agentic Coding187: Agentic Coding
Intro topic: Bad equity vesting schedules
News/Links:- Artemis II
- Gemma 4
- Hatetris
- Dripwarts: School of Drip
- Patrick
- Project Hail Mary
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show- Jason
- FF6 T-Edition
Topic: Agentic Coding- Progression
- Better Code Completion
- Intellisense
- Copilot
- Cowriting
- Cursor
- Roo Code
- Longer Horizon
- Claude Code
- Codex
- Gemini CLI
- Swarm
- Better Code Completion
- Guidelines
- Git
- Planning
- Task decomposition
- AGENTS.md
- Adjusting vs Restarting
- Skepticism
- Future of software engineering?
2 May 2026, 4:59 am - 1 hour 27 minutes186: Becoming a Manager
186: Becoming a Manager
Intro topic: plastic welding kits
News/Links:
- Parse.bot, turn any website into an API
- Gemini 3
- Depth Anything 3
- Wan 2.2 (run on runpod)
Book of the Show
- Patrick
- The Thinking Game (DeepMind documentary)
- Jason
- Plato: The Republic
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick
- Core Keeper
- Pc/Switch/Xbox/Playstation
- Core Keeper
- Jason
- Workers & Resources: Soviet Republic
- PC
- Workers & Resources: Soviet Republic
Topic: Becoming a Manager
- What is a Manager
- Opportunity
- Results + Retention
- Sizing
- Hiring
- Philosophy
- Interviews
- Downsizing
- Hiring
- How to Manage
- Company Goals / OKRs
- Breaking down & claiming company goals.
- Balancing inspirational & practical goals
- Coaching
- One-on-ones
- Career planning
- Performance
- Motivation
- Performance Management Review
- Compensation
- Company Goals / OKRs
- Choosing to become a manager
- Balancing personal and company incentives
- Why Manage
- Mentorship
- Build relationships
- Why to not manage
- Less time for your original joy (coding)
- Less technical influence
- More uncertainty and less closure
- How to transition back to Engineer
- Take the time/energy to get ramped up
- Act as an advisor to your manager
3 February 2026, 1:58 pm - 1 hour 32 minutes185: Workflow Orchestrators
Intro topic: Asymmetric Returns
News/Links:
- NanoChat by Andrej Karpathy
- Pydantic AI
- 1000th Starlink this year
- ChatGPT Apps SDK
Book of the Show
- Patrick
- The Will of the Many by James Islington
- Jason
- Interview with DHH (Founder of Ruby on Rails)
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick
- Factorio
- Jason
Topic: Workflow Orchestrators
- Why
- Batch jobs (embarrassingly parallel)
- Long-running tasks (e.g. transcoding video)
- Checkpointing/resuming
- How
- Message Queues
- Containerization
- Worker Pools & Autoscaling
- History & Backfill
- Steps to run workflows:
- Containerize the workflow definition and send to the cloud
- Containerize all the individual tasks
- Submit job(s)
- Examples
- Airflow
- Legacy but dominant
- Dagster
- Great UX for python developers
- Temporal: https://temporal.io/
- The new hotness
- Ray
- Low-level but very powerful
- Kubeflow
- Designed for ML workflows, integrated dashboard
- Airflow
4 November 2025, 4:00 pm - 1 hour 30 minutes184: Asynchronous Programming
184: Asynchronous Programming
Intro topic: AI Scams
News/Links:
- Coding Adventure: Ray-Tracing Glass and Caustics (Sebastian Lague)
- Boson AI announces Higgs Audio V2
- The Misconception that Almost Stopped AI [How Models Learn Part 1] (Welch Labs)
- A mind-bending conversation with Peter Thiel
Book of the Show
- Patrick
- The Hobbit (JRR Tolkien)
- Jason
- NYT Word Games
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick
- Escape Academy
- Jason
- Multi-modal LLMs to make calendar meetings
Topic: Asynchronous Computing
- What/Why
- Multi-threading in between the lines
- Many of the benefits of multiprocessing without the overhead/complexity
- How
- Coroutines
- Thread-Local Memory
- Blocking vs Non-Blocking operations
- Blocking: arithmetic
- Non-Blocking: Reading from the network card into thread-local memory
- Interpreter locking
- Typescript: Single threaded
- Python: GIL
- Implementations
- Polling (not-Asynchronous)
- Callbacks (interrupts)
- Multithreading (with queues/message passing)
- Promise/Futures
- Async/Await
23 September 2025, 3:00 pm - 1 hour 46 minutes183: Landing a Software Job in 2025
00:00:00 Intro
★ Support this podcast on Patreon ★
00:01:58 Introducing Mark Cunningham
00:07:01 How Do You Find A Job?
00:15:43 How to Get the Best Interview
00:33:06 Tips on How To Pass An Interview
00:38:38 How to Have a Good Interview
00:48:12 What is the Reverse Interview?
00:54:24 What Is The Hiring Manager's Role?
00:57:12 Reverse Interviews: Red Flags
01:14:45 How to Negotiate a Tech Offer
01:23:02 When to Negotiate Your Compensation
01:35:21 Interview Horror Stories
01:39:29 How Do You Deal With Fake AI Applicants?
01:45:32 The Bidding Network31 July 2025, 7:13 pm - 1 hour 37 minutes182: AI Assisted Coding
Intro topic: Getting an entry-level job
News/Links:
- Mario Kart 64 Fully Decompiled
- Q-Learning is not yet scalable
- Grover’s Algorithm
- OrangePi has a RISC-V SBC
Book of the Show
- Patrick
- The Will of the Many (James Islington)
- Jason
- The Intelligence Trap
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick
- Jason
Topic:
- Why
- Speed up development
- Catch errors faster than type checking/compiling
- Writing tedious boilerplate code
- Ask questions and learn local information
- Look good for hiring managers
- How
- Extensions for VSCode & other IDEs for inline suggestions
- Chat with a selection/file
- Command-line Tools run at the root directory
- Local vs Cloud
- Examples
- Copilot (VSCode extension)
- Use the experimental mode
- Cursor (Custom IDE)
- Jumps to suggest changes in other places
- Similar to copilot experimental mode
- RooCode (VSCode extension)
- Copilot (VSCode extension)
★ Support this podcast on Patreon ★30 June 2025, 8:59 pm - 1 hour 46 minutes181: Memory Management
Intro topic: Video Game Prices
News/Links:
- Step one: Jump in the Lava - Abyssoft
- Scalable world models for continuous control
- Clever code is probably the worst code you could write - Engineer’s Codex
- A new, open source text-to-speech model called Dia has arrived to challenge ElevenLabs, OpenAI and more
Book of the Show
- Patrick
- The Muscle Ladder - Jeff Nippard
- Jason
- Metaphysics of War
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick
- Pokemon Trading Card Game Pocket
- Jason
Topic: Memory Management
- Motivation
- Avoid thrashing / crashes
- Allocate resources efficiently
- Keep high uptime
- Where
- OS Level
- Heap management
- Virtual Memory
- Language/Compiler Level
- Cpp
- Garbage collection
- Ownership
- OS Level
- Tools
- Instrumentation
- Export to Datadog / Grafana
- Python: psutil & tracemalloc
- Valgrind
- Instrumentation
- What to do when your program uses too much memory?
- Reduce data sizes
- Compression
- References
- Lazy initializer
- Generators & Back Pressure
- Ring buffers
- Arena allocators
- Disk based caching
- Reduce data sizes
★ Support this podcast on Patreon ★12 May 2025, 3:00 pm - 1 hour 52 minutes180: Reinforcement Learning
Intro topic: Grills
News/Links:
- You can’t call yourself a senior until you’ve worked on a legacy project
- Recraft might be the most powerful AI image platform I’ve ever used — here’s why
- NASA has a list of 10 rules for software development
- AMD Radeon RX 9070 XT performance estimates leaked: 42% to 66% faster than Radeon RX 7900 GRE
Book of the Show
- Patrick:
- The Player of Games (Ian M Banks)
- https://a.co/d/1ZpUhGl (non-affiliate)
- The Player of Games (Ian M Banks)
- Jason:
- Basic Roleplaying Universal Game Engine
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick:
- Pokemon Sword and Shield
- Jason:
- Features and Labels ( https://fal.ai )
Topic: Reinforcement Learning
- Three types of AI
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
- Online vs Offline RL
- Optimization algorithms
- Value optimization
- SARSA
- Q-Learning
- Policy optimization
- Policy Gradients
- Actor-Critic
- Proximal Policy Optimization
- Value optimization
- Value vs Policy Optimization
- Value optimization is more intuitive (Value loss)
- Policy optimization is less intuitive at first (policy gradients)
- Converting values to policies in deep learning is difficult
- Imitation Learning
- Supervised policy learning
- Often used to bootstrap reinforcement learning
- Policy Evaluation
- Propensity scoring versus model-based
- Challenges to training RL model
- Two optimization loops
- Collecting feedback vs updating the model
- Difficult optimization target
- Policy evaluation
- Two optimization loops
- RLHF & GRPO
17 March 2025, 3:00 pm - 1 hour 43 minutes179: Project Planning
Intro topic: Lego event space & retail store: https://www.instagram.com/bambeecave
News/Links:
- StackOverflow Question Count Going Down
- DeepSeek claims its ‘reasoning’ model beats OpenAI’s o1 on certain benchmarks
- Computer Science Papers Every Developer Should Read
- Nvidia Cosmos - an AI platform to change the future of robots and cars - wins Best of CES 2025
Book of the Show
- Patrick: Alice’s Adventures in a differentiable wonderland
- Jason:
- A Beautiful Day in the Neighborhood (Hulu/Netflix/etc)
Patreon Plug https://www.patreon.com/programmingthrowdown?ty=h
Tool of the Show
- Patrick:
- Jason:
- Sqlitedict - Python dictionaries saved to disk
Topic: Project Planning and Management
- Why?
- Gathering feedback
- Identifying risks
- Deciding future headcount
- Documenting / discovering dependencies
- Critical path
- Schedule
- Reduce the bullwhip effect
- How it works
- SMART goals
- specific, measurable, achievable, relevant, and time-bound
- MT is most important
- Gantt Charts
- Scrum
- Agile
- Kanban
- SMART goals
- Tools
- Whiteboard (the generic IRL one)
- Post it notes
- JIRA
- Asana
- OpenProject
- Dealing with uncertainty
- Buffering
- Issues with recursive padding
- Project planning Post-Mortems
- Buffering
3 February 2025, 4:00 pm - More Episodes? Get the App