am.will
@LLMJunky • 32,189 subscribers
StarSwap // Lynk // GooeyPi // DevX Director of n number of agents DevRel @nutrientdocs
Shorts
![Codex update 0.105.0 is out! Despite the fairly pedestrian changelog, this one's a doosie. It's a laundry list of quality of life improvements across the board. - Wispr Voice dictation (hold space to talk) - Theme picker - Codex can prevent sleep on Linux & Windows (I just know there's a joke in there) - Customize Plan Mode reasoning - Many other fixes/updates There's also a complete overhaul to subagents: - New names for readability - Visual display overhaul (way cleaner) - Allow for multi-layered subagent depth (max_depth) - Custom multi-agent role definitions (custom subagents) - /agents now shows both agent names, agent roles, and "dead agents" for auditibility This is the largest single update of Codex I've ever seen! Absolutely massive if you love to use multi-agents. To turn on Voice Transcription, enable: [features] voice_transcription = true Does not work on Linux yet. Well done OpenAI Developers 👏](https://image.24vids.com/tw-2026743879178027416/amplify_video_thumb/2026743822357606400/img/2f-RfwOspQ78vQpu.jpg)
Videos
Exactly 20 years ago today, Steve and Chad gleefully announced that they had sold their 18-month old startup in a $1.65B all stock deal, to Google Today, that startup accounts for ~16% of download traffic, and is worth well over $500 billion dollars. Just last year, they brought in $60B in ad revenue and subscriptions. The YouTube acquisition is certainly one of Google's most consequential decisions, and has truly changed the world. Over 200 billion shorts and >1 billion hours are viewed on YouTube every single day. I wonder if those two imagined it would scale to these heights. I often think of the number of jobs and careers it has created, not just in the building and maintenance of the website itself, but all of the creators that have built businesses around YouTube. It's been estimated that YouTube's creator economy supports at least 1.7 million full-time equivalent jobs across North America, Europe, and India. It's been estimated that the platform pays out ~$25B every single year to creators. In addition to that, YouTube ushered in an educational rennaissance, empowering people to learn and share ideas with ease. YouTube pretty much contains all human knowledge at this point. How amazing is that? So anyway, happy aquiversary and a massive shout out to steve chen and Chad Hurley - thank you for YouTube you absolute legends. 🫡 ▶️
am.will252,785 Aufrufe • vor 9 Tagen
OH MY GOSH ASTRA. GAMING = SOLVED! This is absolutely INSANE! This was not built in Blender. Nor was it built in Godot, Unity, or Unreal Engine. This is just Astra flexing its muscles on the rest of the models to say: I AM THE BEST What can I say? Is RocketLeagueBench saturated? I honestly think so at this point. My jaw is on the floor. This looks so incredibly good. Just you wait to see what people are going to build with this model. WOW.
am.will1,247,832 Aufrufe • vor 1 Monat
I asked Opus 5.5 to write, produce, and create me an educational 'cartoon' about The History of Light - using whatever resources it had available to it. I didn't expect it to make a 36 minute video and spend $200+ dollars though 😂 The study of light is genuinely fascinating, and we have spent thousands of years trying to understand it. This is that story. Opus 5.5 wrote and produced the entire film. I told it to use any keys in my keychain to generate the voices, create any images or video it might need, and to create whatever motion graphics it could to explain concepts in electromagnetism. It used > Gemini 3.8 Flash TTS for the narration, and I must say, it did an excellent job. > Lyrica 3 for the background music, which is also essentially perfect. > GPT Image 2.5, Nano Banana Pro and NB2 Flash for some of the scenes imagery It's certainly not perfect but I was still really impressed with how well it did producing the film. It's very interesting if you're a physics nerd like me. Opus is genuinely creative in a way that I've not really seen with other models. I just gave the model a few constraints and just let it make all the decisions end to end. It even wrote itself credits at the end lol. It's going to be so fun to create bespoke educational videos like this for my children to learn from
am.will158,724 Aufrufe • vor 15 Tagen
NEW: Multi Codex App Run as many unique Codex apps as you want at once, in parallel. Multiple accounts, independent plugins, and customizable dock icons. Add up to 100 Codex apps! No more logging in and out just to swap accounts. Run them all together with ease. Link 👇
am.will49,331 Aufrufe • vor 10 Tagen
I tried this so you don't have to. I know this is going to absolutely shock you but no this does not match the performance of Mythos. A few early thoughts: 1. The limits are pretty bad. I used 100% of my 5-hour usage in less than 1 prompt. 2. I specifically gave it a threejs task because it is an area that SOTA models have made big strides in, that other models just are not great at. I asked it to build a replica of Rocket League. I'll put the prompt in the comments. The game was pretty bad and notably worse than GPT 5.5. Even after multiple fixes, it took 7-8 back and forth with Codex just to get it an almost playable condition. Prior to these fixes, the game was not playable. Maybe it's really strong in other disciplines. I'd love to test that but I hit my limit in 1 prompt lol. GPT 5.5 by contrast did a pretty good job and required no follow ups. Fable would have absolutely nailed this as well. But yeah, early impressions...not great. But I hope I'm wrong. More testing tomorrow.
am.will594,106 Aufrufe • vor 3 Monaten
WOW! I thought Opus 5 would just be a cheaper Fable. I was dead wrong. This model is absolutely BONKERS. I'm genuinely blown away. Opus 5 built the best Rocket League clone i've ever seen, and it did it while consuming only 27% of my 5x Max sub. Sound on. There's layers to this onion. I'm floored rn. At least for building games and 3D modeling, this has to be the best model in the world by a notable margin. Look at the car. The reflections. The details. Playable game & open source code in the comments.
am.will342,609 Aufrufe • vor 2 Monaten
![Look ma new Codex Updates! 0.119.0 and 0.120.0 are here. And with it, a HUGE number of quality of life updates and bug fixes! > Hooks now render in a dedicated live area above the composer. They only persist when they have output, so your terminal stays clean. If you're running PreToolUse or PostToolUse hooks, this is a huge readability win. > Hooks are now available again on Windows > CTRL+O copies the last agent output. Small but clutch when you're pulling a code block into another file or chat. > New statusline option: context usage as a graphical bar instead of a percentage. Easier to glance at mid-session when you're trying to gauge how much runway you have left. > Zellij support is here with no scrollback bugs. If you've been stuck on tmux just because Codex was broken in Zellij, you're free now (shout out Felipe Coury 🦀) > Memory extensions just landed. The consolidation agent can now discover plugin folders under memories_extensions/ and read their instructions.md to learn how to interpret new memory sources. Drop a folder in, give it guidance, and the agent picks it up automatically during summarization. No core code changes needed. This is the first real extension point for Codex's memory system, and it opens the door for third-party memory plugins. > Did you know, you can /rename a thread? But what's really cool about that is, after you rename it, you can resume it with the same name, no more UUIDs. codex resume mynewapp or directly from the TUI: /resume mynewapp > Multi agents v2 got an update to tool descriptions More reliable multi agent environments and inter agent communication > You can now enable TUI notifications whether Codex is in focus or not. Modify this in your config: [tui] notification_condition = "always" > MAJOR overhaul to Codex MCP functionality: 1. Codex Tool Search now works with custom MCP servers, so tools can be searched and deferred instead of all being exposed up front. 2. Custom MCP servers can now trigger elicitations, meaning they can stop and ask for user approval or input mid-flow. 3. MCP tool results now preserve richer metadata, which improves app/UI handoff behavior. 4. Codex can now read MCP resources directly, letting apps return resource URIs that the client can actually open. 5. File params for Codex Apps are smoother: local file paths can be uploaded and remapped automatically. 6. Plugin cache refresh and fallback sync behavior are more reliable, especially for custom and curated plugins. > Composer and chat behavior smoother overall, resize bugs remain though. > Realtime v2 got several significant improvements as well. > You're still reading? What a legend. 🫶 npm i -g @openai/codex to update](https://image.24vids.com/tw-2042844012592189619/ext_tw_video_thumb/2042843968967200768/pu/img/KqqejNVdP4ntsAP6.jpg)
Look ma new Codex Updates! 0.119.0 and 0.120.0 are here. And with it, a HUGE number of quality of life updates and bug fixes! > Hooks now render in a dedicated live area above the composer. They only persist when they have output, so your terminal stays clean. If you're running PreToolUse or PostToolUse hooks, this is a huge readability win. > Hooks are now available again on Windows > CTRL+O copies the last agent output. Small but clutch when you're pulling a code block into another file or chat. > New statusline option: context usage as a graphical bar instead of a percentage. Easier to glance at mid-session when you're trying to gauge how much runway you have left. > Zellij support is here with no scrollback bugs. If you've been stuck on tmux just because Codex was broken in Zellij, you're free now (shout out Felipe Coury 🦀) > Memory extensions just landed. The consolidation agent can now discover plugin folders under memories_extensions/ and read their instructions.md to learn how to interpret new memory sources. Drop a folder in, give it guidance, and the agent picks it up automatically during summarization. No core code changes needed. This is the first real extension point for Codex's memory system, and it opens the door for third-party memory plugins. > Did you know, you can /rename a thread? But what's really cool about that is, after you rename it, you can resume it with the same name, no more UUIDs. codex resume mynewapp or directly from the TUI: /resume mynewapp > Multi agents v2 got an update to tool descriptions More reliable multi agent environments and inter agent communication > You can now enable TUI notifications whether Codex is in focus or not. Modify this in your config: [tui] notification_condition = "always" > MAJOR overhaul to Codex MCP functionality: 1. Codex Tool Search now works with custom MCP servers, so tools can be searched and deferred instead of all being exposed up front. 2. Custom MCP servers can now trigger elicitations, meaning they can stop and ask for user approval or input mid-flow. 3. MCP tool results now preserve richer metadata, which improves app/UI handoff behavior. 4. Codex can now read MCP resources directly, letting apps return resource URIs that the client can actually open. 5. File params for Codex Apps are smoother: local file paths can be uploaded and remapped automatically. 6. Plugin cache refresh and fallback sync behavior are more reliable, especially for custom and curated plugins. > Composer and chat behavior smoother overall, resize bugs remain though. > Realtime v2 got several significant improvements as well. > You're still reading? What a legend. 🫶 npm i -g @openai/codex to update
am.will742,579 Aufrufe • vor 6 Monaten
🥧 Introducing GooeyPi - A GUI for the Pi Family of Agents. Supports: Pi, Oh-My-Pi (OMP), and Prime Agent Very excited to share this with you all. Being into local models, I wanted to have a separate harness for my local models so I don't have to constantly switch configs. And I didn't want to go back to the TUI (nerds). But, I didn't want to miss out on all the features that I know and love. Turns out, I didn't have to. I just needed to build it myself. So I did. Features include: > Fully featured GUI experience > Agentic Browser > Realtime Voice Agent > Voice transcription (local or API) > Computer Use > Automation Desk > Agent to Agent messaging > Ask Question Tool > Git Control > Terminal > Pets > and more I tried to capture all of the most popular features I could in this first pass, but we're only just getting started. Available in BETA on MacOS, Linux, and Windows. I need testers! I have only tested on Mac and Linux. If you want to help me with this project, please reach out. Available:
am.will172,462 Aufrufe • vor 2 Monaten
GAMING = SOLVED (v2) - Opus 5.5 is CRACKED!!! RocketLeagueBench is basically saturated at this point. I received a lot of hate mail from the Rocket League community over the last video I made, but its serious cope. You might think AI created games are bad, or can only look good if they reference existing material, but you're wrong. Even if we say that they're not great now, I think it's pretty clear that we're <12 months away from being able to make games indistinguishable from major studios. All it takes is a bit of time, taste, and prompting and you can honestly build some crazy ish right now. Whether or not something already exists is fairly immaterial. If you can describe it well enough, you can build it. This is by far the best looking Rocket League clone I've ever made. Unbelievable. Opus 5.5 drove this end-to-end, including the cinematic video you see here. The map is insanely good. The Octane meshes and audio were imported, but the rest was entirely generated by Opus (physics based on RocketSim). The best part? It only used 3% of my Weekly usage (x20 plan). It was very slow though. It took about 2 hours for the initial pass, and then about 8-10 prompts to get to this point you see now. Far from perfect but it's only been a day ;) Enjoy!
am.will45,949 Aufrufe • vor 24 Tagen
This new Siri update is absolutely maddening. It's a tale as old as time, Siri being a complete joke. Google has historically just been far better at basically every stage of their lifecycle in terms of voice to text, and phone control. Google Assistant was completely dominant. So you would think considering that Google makes their own frontier models, and the fact they are one of the industry leaders in small models capable of running on Edge devices, that they would easily dominate the mobile space, right? I mean surely that would be true considering Apple has more or less been on the sidelines in the AI race, right? I seriously don't understand how this happened. On device Gemini, even with "Personal Intelligence" really has very little additional utility over the previous deterministic STT "Google Assistant" we've had for the last decade. What are we doing?
am.will206,441 Aufrufe • vor 4 Monaten
THE WAY YOU LOOK AT AI GAME DEVELOPMENT WILL CHANGE AFTER YOU WATCH THIS VIDEO "The AI Slop is Coming for Games." That's right. Introducing... Slopet League Everything you see here was built by Opus 5 and GPT 5.6 Sol in _____🤫. This is not an html page, it is a functional game with a fair bit of polish working inside a real game engine. This has to be some kind of trick. NO. It downloaded a map. NO. Imported assets. Only the Octane meshes, I have other cars that were generated that look just as good. The balls, the map, the effects, the physics, animations, etc were all generated from scratch by Opus 5. GPT 5.6 Sol handled the car meshes. This is only the beginning. I made this in about a day and a half, and most of that time I was waiting on Opus to finish generation. Maybe 25 prompts. I will share my process, and many photos, proofs, evidence of this as we go. This is just a taste, I'm extremely busy this week and working on real life stuff so I'm not going to put any more time into this until next week, but believe me I'm gonna polish this into something beautiful. If you're interested, please share, bookmark, follow, and comment below. I want the world to see just how powerful the language models are TODAY. It's only going to get better from here. Buckle up folks. Oh, and What a save!
am.will131,731 Aufrufe • vor 2 Monaten
The Codex app server was such a brilliant stroke of foresight that really doesn't get enough love Not only are you allowed to use your chatgpt account with any harness, but you can build your own apps directly on top of theirs. They just make building on and with codex such a great experience To demonstrate this utility, I want to highlight the kitty litter app, made by SIGKITTEN. Instead of having to build the entire harness, and all the infrastructure, he's plugged into the app server for a unified experience between mobile and dev machine. When I create a session on my computer, it's automatically available on my phone. All of the chats you see in this video automatically populated when we connected to the app server. All my skills. My agents. My sessions. My folders. My prompts. They're all ready to use - automatically. Because they're exposed by the app server, along with many other endpoints. It's a great ux/dx that really deserves some love. It's almost like they want you to build on top of their products ;) Btw Litter is great 👍
am.will267,261 Aufrufe • vor 6 Monaten
Nice. Cursor just dropped their new "Glass" alpha, and they're leaning heavily into the simplified coding GUI trend that's been blowing up lately. First impressions are really positive. And just look at how insanely fast Composer 2 is. First impressions? Drop yours 👇
am.will217,789 Aufrufe • vor 7 Monaten
ICYMI PSA Upgrading your Anthropic account = a full weekly reset. It's actually amazing. And quite generous. Video evidence below. Yes this works for $20 -> $100 and $100 -> $200 I dont see them removing this any time soon because it's actually quite difficult to abuse. You'd first have to downgrade, wait a month, and then upgrade again.
am.will26,915 Aufrufe • vor 21 Tagen
anything under deepseek v4 flash basically doesn't exist to me. with <200GB of vram, i'm getting ~800K context @ 200 tokens per second. It's not quite Cerebras fast, but damn. It is a joy to use, and its basically somewhere in the opus 4.7/4.8 range. It's incredibly good at browser control too. this model is a beast. w/ qwen 3.8 27B being multimodal though, it might just be worth switching to, especially if you want to run a lot of parallel agents will 300toks/s be possible? Pretty addicting let me tell you. Imagine a computer use model this fast. watch below, this is basically human speed.
am.will72,729 Aufrufe • vor 2 Monaten