Darshan

99 posts

Darshan

Darshan

@THEdort01

22 | 🛠️ building context-engine (coming soon...) | Aspiring AI Engineer | Prev @ Layer5, LFX mentee 25' | Building AI tools | agents & DevTools

Earth Katılım Eylül 2019
365 Takip Edilen39 Takipçiler
Darshan
Darshan@THEdort01·
@thekitze Haha, I second this. what do you do while your agent's working tho?
English
0
0
1
32
kitze
kitze@thekitze·
i blocked twitter/reddit/youtube on my work computer because i'm so fucking tired of the "tell agent to do something" → "go and scroll/watch something" loop. it's mentally exhausting
English
11
0
27
1.1K
Darshan
Darshan@THEdort01·
My main point is Obsidian’s visual graph view. It offers a rich, bird’s-eye context of your entire knowledge base—showing connections, clusters, isolated notes, workflow structure, gaps, and broken links at a glance, making navigation far more intuitive. When we’re all coding in a rush and wasting time re-exploring the codebase, this could genuinely be a game-changer imo.
English
0
0
1
15
Darshan
Darshan@THEdort01·
Obsidian plugin with full two-way sync. Codex could pull context from my vault AND write new notes or update existing ones. The graph view would be insanely useful: visualize codebase architecture, show how new features connect to existing modules, highlight knowledge gaps in my personal wiki, or even map dependency graphs and idea relationships automatically. Basically: turn my Obsidian graph into a living map of both my code and my thinking.
English
3
0
12
516
Romain Huet
Romain Huet@romainhuet·
We’re thinking about the next wave of Codex plugins. What’s one you’re missing today?
English
218
8
272
16.5K
Nicolas Zullo
Nicolas Zullo@NicolasZu·
I have feeling Codex Reset incoming We will hit 5M Codex users on 5/5 at 5:55 thanks to GPT-5.5.
English
38
8
724
33.5K
Darshan
Darshan@THEdort01·
@theo Whats the prompt my guy???
English
1
0
1
5.8K
Theo - t3.gg
Theo - t3.gg@theo·
I sent a single message on Copilot and it did over 60m tokens. It's still going. $30 of inference so far. In their current billing model, you get 1,500 messages, regardless of how expensive each is. I'm pretty sure I can do $45,000 of messaging on this plan
Theo - t3.gg tweet media
English
184
23
2.9K
228.6K
Darshan
Darshan@THEdort01·
@haider1 Antigravity as well, with its infamous “Agent terminated due to error” msg
English
0
0
2
231
Haider.
Haider.@haider1·
gemini 3.1 is still an incredible model but i've had enough of its hallucinations because when it gets things wrong, it can suddenly feel like using a pre-gpt-3.5-era model, even with a short context really hoping for an updated flash model and gemini 3.5 pro this month the ".5" series is usually better
English
44
10
350
18.8K
Darshan
Darshan@THEdort01·
@cjzafir Why is kimi below claude and GEMINI!?😭😭😭 is this ragebait?
English
3
0
12
5.8K
CJ Zafir
CJ Zafir@cjzafir·
Best models right now: 1st > GPT 5.5 (extra high) 2nd > Deepseek v4 pro (think max) 3rd > GPT 5.4 (extra high) 4th > Claude Opus 4.6 (extra high) 5th > Gemini 3.1 pro (thinking) 6th > Kimi 2.6 (thinking) 7th > GLM 5.1 (thinking) 8th > MiMo v2.5 pro 9th > Qwen 3.6 max (thinking) 10th > Claude Sonnet 4.6 (high) PS: Yes no Opus 4.7 because it's very unstable, unpredictable and nerfed model right now.
English
98
79
1.4K
118.1K
Darshan
Darshan@THEdort01·
@sama The limits are consumed ridiculously fast with xhigh reasoning, almost as if the pro subscription is like a paid preview option with 2-3 prompt allowance
English
0
0
2
1.5K
Sam Altman
Sam Altman@sama·
5.5 xhigh in fast mode is really good i think i got psyoped by twitter on medium for a bit
English
751
110
6K
577.5K
Darshan
Darshan@THEdort01·
@replit down already 😅 Replit Agent has been surprisingly solid for vibe coding — fast iteration, infra handled, and fewer setup loops. waiting for the issue to be fixed...
Darshan tweet media
English
1
0
3
192
Darshan
Darshan@THEdort01·
@levelsio Essentially with every step forward we think we’re going, we go two steps backwards unknowingly in modern technology…
English
0
0
2
343
@levelsio
@levelsio@levelsio·
Yep we're essentially all gassing ourselves with CO2 especially with the new energy saving insulation that makes houses air tight boxes, very bad for our brains!
Johannes Schmidt@spaceMonster

@levelsio Western civilisation is sleeping on this problem since it only was created 20-40 years ago. Before that windows where not that airtight. Also cheap CO2 sensors ate only available for 10 years before that they where prohibitly expensive.

English
53
4
348
50.2K
Darshan
Darshan@THEdort01·
@elonmusk Grok Voice @ 67.3% on Starlink — on-device or hybrid inference? How much score survives real satcom latency/jitter?
English
1
0
7
518
Darshan
Darshan@THEdort01·
@tycoi2005 @pseudokid Ahh got it! My OpenCode desktop updated fine, but the DeepSeek model I was using in the CLI was stuck in Warp on the old version.
English
0
0
1
64
raymel 👋
raymel 👋@pseudokid·
OpenCode Go is practically unlimited with Deepseek V4 Flash Even at max thinking, which is pretty decent, I'm spending ~2% of my monthly limits a day At that pace, it's ~50 8h session days a month! Double that to 16h a day, and it gives you 33.3 days making ~4%/day its ceiling
Lotto@LottoLabs

Update on Opencode Go It’s great value for $5/month, there’s really no reason not to do the first month. At $10/month it’s still good value and gets you access to all sota OS models. You can’t daily drive it without hitting limits on the big models but w/ Kimi x3 you won’t hit limits unless you’re insane. Overall highly recommend the first month, then make your own decision.

English
39
20
676
53.3K
Darshan
Darshan@THEdort01·
@theo Truth shouldn’t cost a million bucks, inference being this slow in 2026 is just not good enough, vibercoders are watching, fix it or we’re all dipping...
English
0
0
3
3.1K
Darshan
Darshan@THEdort01·
Wow, OpenAI just gifted a full month of ChatGPT Plus completely free! 🎉 Thank you @OpenAI @ChatGPT @sama and the team — this made my day! Gonna buildmogg my way through goblin-season™️ with god-tier write context.
Darshan tweet media
English
0
0
2
179