Talked to my barber today during a haircut. Told me they're not happy with alteg.io & they're looking for a custom solution. I'm now preparing a demo for my next visit covering all their painpoints: internationalization, custom integrations, customizations, etc
The useful part:
I can forward a Telegram video to my agent and it downloads it, transcribes it, and sends the result back automatically. Lets me analyze my video content and see which content type creates most engagement
Built a Hermes agent on @Railway with a @Tailscale forward template. Connected it to @telegram and had it download and transcribe videos using @Zai_org GLM 5.3 Flash
Built a Hermes agent on @Railway with a @Tailscale forward template. Connected it to @telegram and had it download and transcribe videos using @Zai_org GLM 5.3 Flash
I've made it open source github.com/ramziddin/persona…
I also love that @Railway now lets you do infrastructure as code
At this point I start giving AI the tasks in the To Do column while I have it do my current in Progress Tasks
i really miss having a terminal pane in claude code desktop, similar to what codex has
@claudeai please!
I feel more disconnected to the people I follow now
Final request to @nikitabier (for today :D)
If last post set to or posted as "Accounts you follow and who they follow can reply"
Remember the setting and set is as default for any new post (unless we then change it of course)
For those wanting to use Claude Cowork's Dispatch, don't forget to run `caffeinate` on your mac, it'll prevent your mac from sleeping
I used to watch PewDiePie play Happy Wheels when I was 13, now I’m 25 and watching him train his own AI models
PewDiePie trained a frontier model at home and beat OpenAI and Gemini
> be me PewDiePie
> play games on YouTube and scream 24/7
> become a meme reviewer
> get unfathomably famous
> "fuck that" I'm a family guy now
> retire
> move to Japan with my beautiful wife
> mfw I'm a dad now
> "as a dad I must do dad things"
> scratch that
> "as a dad I must do frontier AI research"
> goal is to beat GPT-4o at coding (16% on Aider)
> buys $30000 GPU setup
> reads DeepSeek paper
> decides to start massive GitHub scraping and data augmenting run
> not good enough
> read Magicoder paper
> generate tons of synthetic coding data
> train a new model
> guuuuuh. the data made the model worse
> mfw I just wasted months for nothing
> decides to lock in and try again
> makes model worse again ffs
> try again
> finally beating GPT-4o on data (16.1%)
> not satisfied
> "I should simply train a reasoning model"
> reads more papers
> start experimenting with more synthetic data
> "Mhhh something doesn't smell right"
> house almost burned down due to power connector
> shrug
> just buy a new one
> mfw computer is now crashing 24/7 generating synthetic data
> new plan: just call DeepSeek API for high quality synthetic data
> train model again
> 17.2%
> performance fluctuates slightly on each eval run
> big brain idea: repeat eval until we randomly reach >18%
> sike actually got 19.6%
> feelsgoodman.png
> nvm the benchmark was contaminated and I was training the wrong base model the whole time
> rerun everything again
> new score: 4.4%
> you read that right REEEEEEEE
> almost get a heart attack
> "have you tried plugging the device off and back on again?"
> change nothing and just retrain again
> 25.3 %
> LETS FUCKING GOOO
> realize that 1/3rd of the benchmark was not running. guuuh
> scared shitless it would score below 10% again
> run yet again. the whole thing this time
> Thirty fucking six percent
> accidentally beat Gemini 2.0 Pro Exp and GPT-4.1 mini
> pops the AI bubble
> "I want moaaaar"
> finds some more post-training data
> 39% babyyyy
> realize at the end that I was just benchmaxxing Aider polyglot
> next quest: run SWE-Bench and other coding benchmarks
> "I failed a thousand times, but prevailed in the end"
> just a little sad side-quest
> probably going to train GPT-6 myself by next month
i got tired of ai writing slop code, so I taught it solid principles
open sourced an agent skill that enforces tdd, clean architecture and proper design patterns
𝚗𝚙𝚡 𝚜𝚔𝚒𝚕𝚕𝚜 𝚊𝚍𝚍 𝚛𝚊𝚖𝚣𝚒𝚍𝚍𝚒𝚗/𝚜𝚘𝚕𝚒𝚍-𝚜𝚔𝚒𝚕𝚕𝚜
works with claude code and other agents
details: github.com/ramziddin/solid-s…
"Ok Claude, inject 5 grams of creatine via creatine MCP, launch gym session workflow (use gym-skill) and make my side project generate $10k/mo. Make no mistakes"
Hey @bcherny, why does Claude randomly stops showing the "Clear context and auto-accept edits" option with Plan Mode? I don't understand why it does that and how to show it
Codex gives Claude a task from interview_tasks.md with tasks ranging in difficulty levels, makes Claude load the skill and then...
