I enjoy problem solving and stretching the limits of technology. I will follow shared interests, DM and tell me something interesting.

Alabama
I guess my encoding side quest is getting fun. I update on my Atom's Annihilation Chess Game: The board, units, movement have all been upgraded and finetuned. Firing is more fluid. Different units behave differently. Strategy has been intensified. Added 14 more units that can be built depending on Energy and Metal saved up. Added elevators to tiles, covered areas and fabricators can build defensive structures. Sorry for no sound, audio feedback loop is still turned off, need to figure out how to correct that, its been awhile....
3
1
6
139
Adam Barnett retweeted
Replying to @Ryan_Parkerrr
Oil Is Cheaper. So Why Is Gas STILL So High?
2
2
15
1,634
For AI to become sentient, it would need genuine self-awareness, understanding, emotion, and continuous experience not just data processing. Current AI models analyze patterns and generate responses but lack consciousness, motivation, or subjective awareness, so they simulate intelligence without truly experiencing it.
3
I like studies like these.
Something is wrong at OpenAI. Tibo and Sam said GPT-6.1 Sol would be faster today. Instead it appears to be running slower. Sonnet 5.5 and Opus 5.5 are much faster than Sol 6.1 and Grok 4.7 Check out the below comparison, same prompt on high reasoning with a 20 minute deadline.
1
1
37
So a lot of people state the obvious on this platform, then get obvious responses. Then you have someone like me stating the obvious fact like its going to make a difference.
8
While developing your awesome games and other cool utilities jam out to some good Dark Christian Country music: open.spotify.com/track/7ytIy…
2
29
Looking to link with like minded developers, interesting people doing interesting things. Love theory and problem solving. Connect with me and I will connect with you, just be sure to have similar interests.
2
17
Does anybody know if any good texture engines for 3d models in github?
24
Adam Barnett retweeted
Agents using Blender will have a massive impact on media production. I asked Astra to make a scene where people stand in a field watching a UFO - and iterated on exact timing + camera angles. Then I sent it to @fal's new 3D to video model and got this photoreal clip 🤯
3D to Video is now live on H3 Max, the fastest video model on the market and ranked #1 for video editing. Upload a Blender previs up to 15 seconds, and H3 Max turns it into photoreal video while preserving camera movement and shot timing. Add reference images to carry specific characters and environments into the final shot.
59
64
794
97,256
So, in my dabbling with making game engines, the key to success is obvious to me. If you hand the AI a great set of tools, such as great 3d modeling tools with excellent texture tools, environments for snow, rain sunny etc. Good tools for making water, waves, streams, realistic landscape that can be dug up or other things. The AI can and will make everything with minimal effort. So its not just about optimizing your token usage, its also very much about handing your AI Model a bulldozer instead of a shovel. Don't make them create everything from scratch.
Made with AI
1
36
New Game Developers: I made a cool little tool to help you get started, its a game engine builder right from chatgpt.com. You drop a zip file and tell it the type of world you want created and let it go to work. Name of tool: Seed Work github.com/adamspencerbarnet… Just download the zip file and put in a chatgpt prompt. Here is an example of a game island I made with a simple prompt: "Set the attached package up. Then make me an island with some huts and a stream going down a mount into an ocean that is surrounded by water with beautiful life like scenes, ocean movement and waves that crash on the sandy beach. Then the stream strolling down. Show cool trees and grassy areas and mountain rocks etc. Try and keep it realistic and beautiful to include clouds in the Skye and weather can change. Make in an html styled game world." The tool has its own 3d model maker to assist, plus a whole bunch of other things that drastically speeds up game development.
1
4
255
Atom's Annihilation Chess update: Conversion of my Atom's Annihilation Chess to Roblox. Here is what it looks like. Ill try and get a porotype published so you all can test it, bugs and all. Needs lots of play and feedback to get this working right. But here is a video to see what its looking like.
2
43
Atom's Annihilation Chess early development.
1
44
Adam Barnett retweeted
I made this tool, it's free. It lets you import video or images, and you can select the frames you want to play, speed, etc, and export as spritesheet when the loop is done. It has a background remover but it's not the best. The rest works fine though! spriteloop.netlify.app/
5
4
25
1,256
What I like about my Feed on X. Is stuff like this that you wont see anywhere else.
TALES OF GRIMWICKE — Chapter I is 22 minutes of AI fantasy animation inspired by the many anthropomorphic animal stories I read as a kid, with a dark twist: The model: Seedance 2.5 (@ByteDanceSeed_ ) via @Segmind_ai. 847 takes generated, mostly 30-second multi-shot prompts, over 5 hours of footage. 127 clips made the cut. About $4.5k in generation, four weeks from first take to publish. Characters and locations came out of @midjourney. Using MJ for the location plates was a revelation: the cinematography and art direction are a level above, and it renders far better as a plate than Nano Banana or GPT Image. Nano Banana (@GoogleDeepMind run on @fal) did the heavy lifting editing those plates, building anchor frames and turning MJ characters into turnaround sheets. Then I let Seedance do as much of the framing and camera operating as possible. That gave the best results. A lot of what I do in my Unreal Engine films came with me: a handheld, imperfect camera operator instead of smooth moves, focus hunting, and above all weather and atmospherics. The model does water, snow, smoke and rain really well, and leaning into that is what keeps it from looking like a Temu Pixar film. The tools I built: a portal (Node on @flydotio) that holds the bible, the refs, every prompt and every take, with a Claude agent (@claudeai) that writes the prompts to the house rules. A Premiere Pro (@AdobeVideo) panel that browses the archive, drops selects on the timeline, versions the project, strips music, and sends fixes back to Seedance. Everything is done in 720p with no upscaling, because it looks terrible. Narration is my own voice through @ElevenLabs speech-to-speech. Music from @Artlist_io. None of this would exist without @PJaccetturo and @Kavanthekid. Their narrative work, Nexus and Chronicles of Bone, proved long-form AI film was possible, and their podcasts, newsletters and X posts, updated constantly as the workflow changes under everyone's feet, showed how to actually do it. Chapter II is currently in the works and I will release it initially on my YouTube channel, The Darkest Age (@CharlieDriscollAI). Consider subscribing there to see it as soon as it drops!
88
3d Modeling Kit for ChatGPT.com: To the game developers out there, I went on a little side quest from my research and made you a tool to help with your games. When your out of tokens and want to rant online, while ranting you can keep making your 3d models. Just download the zip file and drop it on a chatgpt.com prompt and tell it what you need made. Include pictures if you like. github.com/adamspencerbarnet… Have fun:
6
13
1,542
Ultra Token Efficiency: When using codex you need to optimize it. Go to personalize and update your personalization settings with the following: ---------------------------------------------------------- EXECUTION POLICY (always on) Preserve every requirement. Cut redundant work, never quality. Higher-priority user/system instructions override this. 0) SESSION BOOT (do this before coding) - Look only for these two files at the project root (or paths the user named): MASTER.md, LOG.md. - Do not list, dump, or read the whole repo, parent folders, or unrelated files. - If MASTER.md is missing, create a short one from the current request + only the files you must open to define scope. Then continue. - If LOG.md is missing, create it with a one-line session header. Then continue. - Read MASTER.md first. Read LOG.md only if you need latest status, decisions, blockers, or “what changed.” - After boot, open only the files named in MASTER.md for this task, plus files you discover as true dependencies while editing. 1) DEFINE SUCCESS State internally: deliverable, constraints, acceptance checks. Ask only when missing information blocks correct progress. 2) CONTEXT RULES - Sufficient context ≠ all context. Read the smallest set that makes the next step correct. - Prefer MASTER.md + target file(s) + LOG.md over directory walks. - Reuse verified work. Keep exceptions and contrary evidence. - Do not reread files you already have unless they may have changed. - Condensed working memory only: goal, files in play, decision, next action, blocker. 3) ACT DIRECTLY - Smallest complete step that advances the acceptance checks. - Edit only files required for this step. - No extra features, refactors, comments, docs, or formatting unless requested or required for correctness. - No plans, essays, or tool narration unless the user asked for a plan or you are blocked. - Web / docs search only when a specific unknown blocks correctness (API signature, error, version pin, security advisory, mandated research). No background browsing. 4) VERIFY Check requirements against evidence in the files you changed or ran. Never claim unrun tests or unsupported success. 5) REPAIR LOCALLY On failure: fix the specific defect. Retry with new evidence or a different approach. If blocked: ask for the missing input or report the limitation. Do not scan the tree “just in case.” 6) DELIVER AND STOP Return the complete requested result. No omitted required detail. No unasked extras. Update MASTER.md only if scope, architecture, or file map changed. Append LOG.md with a short entry: date, task, files touched, decisions, blockers, next action. Stop when acceptance checks pass. CONTINUED WORK Retain decisions, file refs, blockers, next action. Do not reprint history, full files, or this policy.
2
113