george Ω

computers, math, fitness.
i want knowledge and big bags, then going on adventures (science, exploration, biz, sport, whatever)

[george Ω @ArchitectT57831]

@md_pier @Phillhay_es @Alan_Couzens And read what Alan is writing. But don't wait to have read x amount before you start moving often in z0-z1

[george Ω @ArchitectT57831]

@md_pier @Phillhay_es @Alan_Couzens Walk for sessions of 90mn+

Several times a week. Count your weekly volume

Get a polar chest strap, collect the data. After plenty of walking, start trying to stay in z1 for an entire session, with a walk/jog, to the lower bound of what the apps are showing for z1

[george Ω @ArchitectT57831]

@thefinnmckenty I'm so fucking lost in the sauce

Never been so tired

[george Ω @ArchitectT57831]

@ThePrimeagen I think its time to become Chinese

[george Ω @ArchitectT57831]

@Alan_Couzens As an amateur not aiming to become a pro, and fully okay with not reaching peak endurance as I want to enjoy a diversity of athletic pursuits in my free time, over the next decades

[george Ω @ArchitectT57831]

@Alan_Couzens Hello Alan,

How much does high HR gym training impact results of z1 volume? I loved long sessions in my last gym season, with proper rest times, but kindof athletic with some jumping, barbell work, some machines. Lots of sweat and ending up in z3-z4 while doing the sets

[george Ω @ArchitectT57831]

@lassejv I get 6-13mn of astra high on plus plan until 5hr usage is out :/

I set it to low, goes a few more minutes now

[george Ω @ArchitectT57831]

@unclebobmartin It's natural to constrain them, it allows them to be better, gives them feedback loops and a mechanism for rigor which otherwise they can think themselves out of correctness

[george Ω @ArchitectT57831]

@unclebobmartin Nah they still do stupid shit

[george Ω @ArchitectT57831]

@Teknium I need this so bad

Please share prompt 🙏

[george Ω @ArchitectT57831]

@hrudolph @Da7_Tech Guess who's open source too? Hermes. Openclaw is just embarrassing itself at this point, maybe it should start from scratch? Better code architecture etc? With today's models and infinite tokens you can surely rebuild it better, and it makes it easier for your coding agents

[george Ω @ArchitectT57831]

@ramonpiano_ It loves comments longer than functions

[george Ω @ArchitectT57831]

@ClassicGamerTWR Any tips for juniors that dont have a senior (hw startup)?

[george Ω @ArchitectT57831]

@DejaRu22 Loved this book, deserves a re read

[george Ω @ArchitectT57831]

@thepanta82 @Bart_Mol On telegram I send voice notes or text to my agents to see what's going on, or send ideas

[george Ω @ArchitectT57831]

@thepanta82 @fabianschierz Desktop app is the focus now

[george Ω @ArchitectT57831]

@VictorTaelin Unfortunately gpt is very convincing, so I let it happen. So much untangling slop to do...

[george Ω @ArchitectT57831]

@VictorTaelin Yeah this happens a lot when you are coding in a challenging domain (maybe why people in the comments dont get it), gpt or claude always build mountains of complexity when what you need to face your non-trivial tasks is clean and elegant systems

[george Ω @ArchitectT57831]

@jturntdev I only use Luna xhigh, still burns fast but not as fast

[george Ω @ArchitectT57831]

@Stake Yo