tristan
624 posts
tristan
@tris_does_stuff
Stuff doesn't do itself | https://t.co/iY7RVi1AC0
Germany Katılım Eylül 2025
337 Takip Edilen59 Takipçiler
“It is possible there are some hiccups soon”
Oh no
Sam Altman@sama
5.6 sol growth is insane. the inference team has done heroic work to be able to support demand. we are going to move mountains to continue to scale, but it is possible there are some hiccups soon.
English
@thsottiaux Surely not a new product called Codex. Surely not.
English
I had double jaw surgery 3 weeks ago
I woke up and could finally breathe through my nose for the first time in my adult life
I no longer have: obstructive apnea when sleeping on my back, bruxism, TMJ pain, or severe dental malloclusion
My lower jaw is no longer giga-recessed
I am mewing basically 100% of the time without even thinking about it. This was previously impossible. My tongue just naturally rests on my pallet. Proper tongue posture is now effortless
Furthermore, my head no longer naturally drifts forward. My posture drastically improved, literally over night.
The first week was brutal, but things are slowly improving
I am still super swollen and look a bit like a chipmunk, although swelling has come down a lot. The final result can take 6-12 months to eventually show
Currently on liquid food. Will be unable to eat a steak for several months
But it was 100% worth it
English
this app does $800K MRR with 200K monthly downloads…
they’re using every angle possible, and it works
indirect videos on tiktok at scale (about insecurities or feelings) makes people comment
you reply to those comments, instant warm lead conversion for the user that commented and the people who are going to view the comment (more people view comments than you think trust me)
now sometimes tiktok videos take off later than usual so comments are viewed even more
running search ads on your keyword also helps a lot when a user goes from the tiktok video to app store (IF you aren’t running tt spark ads and your account is lesser than 1k followers)
comment ‘SECURE’ if you want the full playbook (must be following to dm)
English
Clipmem 0.6.0 is here!
• Better search with honest match scores
• Safer restores with rollback and no duplicate captures
• Consistent privacy rules across every capture path
• Source-safe image previews and durable OCR jobs
• Stronger deletion and database integrity checks
• Just in general better
Save all your clipboard history to a local SQLite db (redacting secret-looking-things if you choose). Images get OCR'd by a local model so they're searchable too, and there's a nice CLI for your agent and a nice MacOS GUI for you :)
English
English
@tris_does_stuff @theo @risingtidesdev @TraycerAI thats actually helpful, thanks for the feedback, we're always open to hearing out the users!
after reading your idea, how about something like, @ mention inside agent guide for easily mentioning provider or model name along with reasoning.
English
Ah just the agent setup being a prompt, which I can see the appeal of but sometimes I just want some dropdown boxes 😅
I enjoyed it in the end, typing out my heuristics for which models to choose when based on my own internal vibes.
Maybe like 2 options, one is the prompt based method, the other is dropdowns like
UI agent -> Opus 4.8 high
Exploration agent -> gpt-5.6-luna xhigh
But saying this now I realised I could have literally typed this into the box, instead of the long version I went with 😂
You guys are doing great don't listen to me
English
@tris_does_stuff @theo @risingtidesdev @TraycerAI what felt like setup? 👀
We’re trying to make everything feel magiccc
English
@theo @risingtidesdev Same, I thought it was just me. I was excited to use Fable with it with GPT-5.5 but froze every time I tried it.
I've been using the @TraycerAI app, takes a little setup but the subagent abstraction with agent to agent mail is genuinely one of the best I've used.
English
@risingtidesdev It froze my terminal on install for 2 minutes bro. It's slop.
English
Just set it to “high” and stop worrying so much. The model already adapts thinking levels really well.
Haz@diegohaz
Low, medium, high, x-high, max? I wish I didn’t have to think about model effort levels. Do they recommend starting with lower levels just so you can save tokens, or is the output actually better? We often don’t know how much effort a task will need upfront. We figure that out as we go. Should we then keep monitoring the agent’s work so we can set the right effort level mid-task? This sounds suboptimal. If that’s the case, shouldn’t the model automatically determine mid-task whether it needs a higher effort level and adapt automatically? From experience, I can tell you that even frontier models are terrible at measuring effort at the beginning of a task.
English
For agentic coding, one can say:
- Unless you need Terra Ultra perf, it's always better to use a Luna model with higher effort setting (same or better performance but cheaper).
- Forget everything below Sol High, use Luna with higher effort settings here
- Forget Sol Extra High, use Terra Ultra here
- The extra cost of Sol Ultra is probably not worth it over Max
English
@eigenrobot 21 years old for a cat is awesome. Hope you're doing alright
English
21yo cat's not eating :/
nibbles kitten food when offered but barely. not touching milk, cream, tuna, butter, deli meat
feels bad man
English
@shipordie_ @justraud @daniel_a_jm @bugsntea @dimsav @TravelingTice @kylekane @CodeWithStu Who's that loser at #1
English
Top builders by commits in the last 24 hours:
🥇 @tris_does_stuff — 95 commits
🥈 @justraud — 51 commits
🥉 @daniel_a_jm — 32 commits
4. @bugsntea — 31 commits
5. @dimsav — 30 commits
6. @travelingtice — 22 commits
7. @kylekane — 21 commits
8. @codewithstu — 21 commits
9. @themattlok — 18 commits
10. @fortisaleo — 17 commits
Next board drops tomorrow at 2 PM UTC.
English
The new @TraycerAI desktop app is legit, works very well for Fable orchestrating Codex/Opus sub-agents. Probably the smoothest experience with sub-agents I've had so far.
English
Yes. Besides the fact it is obvious to anyone who bothers paying attention to women, there is also plenty of research on this. Women prefer more rugged men on lead up to ovulation, and doughier lads during the lead up to menstration. This is only one of many such patterns.
The amount of difference changes from woman to woman, as well as the specifics of her tastes, but the tastes do always change at least a little.
Ideally a husband is malleable and attentive enough to act as every aspect of man his wife desires throughout the month.
English
Literally no idea what you're trying to say.
"looks like he trains for hypertrophy and probably isn't particularly strong"
> Shows someone who trains for hypertrophy and is stronger than most males.
"That's what I was saying, training for hypertrophy is no different from training for strength"
Can you please write more coherently?
English
@tris_does_stuff @VertexFan My ninja, that's awesome, but you just made my point. You "train exclusively for hypertrophy"? How is that different from training from strength then? Oh wait, that was what I was saying
English
Women think there are two different kinds of muscle, and one is for show
Daisy@h3ll0again
we like a big bulky strong man that xan handle shit, not a vain bozo who very likely isnt functionally strong because he lifts for aesthetics
English