Conviction1000% (@ConvictionFAQ) header
Conviction1000% (@ConvictionFAQ) profile picture

Community Lead @MindoAI / Hermesmaxxer

Joined February 2025
Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)Photo by Conviction1000% (@ConvictionFAQ)
How do you grow the community from 952 to 59,609 in 3 days? Here is how: = Mindo AI delivers real value to the community. = Mindo AI is always on, it never sleeps. = Mindo AI is still early. In Mindo AI we trust! Let’s look at the growth of Mindo AI’s Discord. @MindoAI
69
17
3
201
9,410
FABLE 5.2 NUCLEAR EXPLOSION >1 SHOT >30 MIN >PS 2 STYLE + MAX REASONING
8
202
AGENT vs SUBagent when to use each one Here is how I see it: >Agent: one loop, sees every step >Subagent: a fresh helper, own context and tools >Never sees your chat, brings back only a summary Same model, very different bill. So when do I reach for subagents? Whenever the work splits into pieces: - research and search running in parallel - keeping my main context clean - a cheaper model for each small job (Hermes routes one per subagent now) 50 SUBAGENTS FOR SIMPLE QUERIES (bruh) Google tested 180 agent setups and parallel finance tasks got 81% better with a lead plus helpers. Anthropic got 90.2% better research with Opus 4 leading Sonnet 4 subagents vs Opus 4 alone. And when do I stay with one agent? When every step depends on the last one: - planning and step by step builds (multi-agent setups did 39-70% worse there) - routine and scheduled jobs, same steps every day - anything where two helpers can disagree (in Cognition's Flappy Bird example one drew a Super Mario background) Also it's not cheap. One agent burns ~4x the tokens of a chat, a lead with subagents ~15x. Anthropic's early agents even spawned [Subagents read. The main agent writes.]
1
8
210
OPENAI SHUT DOWN PRISM 2027 18 years old Physics student \ quant (me upset)
3
118
Skip 3 days per week now 20 Parallel agents can do now Big deal with 2x larger limits on Codex I really think people don't appreaciate it enough You can voice your tasks as 1 hour speech to Codex And 20 agents will start executing it.
3
77
The logic is following >Astra is settled for subagent work >Sol for not such demanding tasks >Astra better for parallel tasks
In the new unified ChatGPT, if you want to use the Pro model, you have to select GPT-6 Sol instead of Astra and then set it to the highest reasoning effort, which is GPT-6 Astra Pro Selecting GPT-6 Astra gives you access to Ultra, not Pro
3
86
I'm thrilled to announce I've committed to open-source project I mean literally committed How do you like LinkedIn style hook?
4
73
Except awful annoying miserable Halloween sound I love Discord as Community Lead of 60k server
Discord is about to put your public messages on Google Discord will start making messages from Discoverable public servers searchable on Google and other search engines from mid-December. The feature is ON by default in most regions, but you can opt out: User Settings → Data & Privacy → turn off “Show my message content from Discoverable channels on web pages.” In the EU, UK, Canada, Brazil, Australia and South Korea, it will be OFF by default. Private servers, DMs and group DMs are not included. Usernames and avatars won’t appear alongside indexed messages.
2
90
Anyone who runs agents see what yours actually did >You open the session log >Wall of JSON >You scroll >You miss the one error that matters I don't want to read Cua Driver logs. I want to see the session. So I drew one real session as a manga: -Agent: Codex + Cua Driver -Task: X article -> folder -Calls:41 -Time 19 min 41 s What you catch in 13 seconds: what it clicked(in the background) what it could touch my Chrome, only after --grant where it got stopped -> unbound JS refused REAL LOGS, NOT MOCKUPS Session Replay every step from the log is a panel, every error is a panel too. I want it inside Cua: run a session, get your replay (design not final) @trycua Do you think it is a good way to read reports of your sessions?
3
102
You are not going to know which one is GPT 6.1 Terra Put myself in a dirt for these ones >Both generated from Opus 5.5 Max prompt >Different harness >You have 0% chances to spot So have you cancelled your Anthropic sub already?
66
Fable 5.5
i could probably just tweet "fable 5.5" and get insane reach lol
1
4
611
Kimi K3.1 Pro just dropped and talks 2.8x less sh*t - 3T params - 1M context - Open weights TerminalBench 4.0: Opus 5.5: 66.4% Kimi K3.1 Pro: 65.7% GPT-6 Astra: 58.2% Full weights coming by Nov 4 Open-weight model 0.7 points from Opus rn I'll imrpove my larp engine with it
1
4
93
12x agent code pushed AI to Rust AI Engineering was Python glue around an API call. Anthropic merges a day: May: 2,500 September: 30,000+ Human scale infra broke first. CHECK PIP Look who is rebuilding it in Rust: - LayerFS by Ephemeral AI Lab: forkable file system per agent, since shared folders create races - AMD: small elite team to push Rust deep into the GPU stack. Job post says "Rust is not incidental to this role" why Rust? nobody can read 30,000 changes a day by hand. So [the compiler] checks first: - two threads change the same data at once: won't compile - code uses memory that's already freed: won't compile - hidden crash bugs: Rust tools check for them before release In Python these bugs reach prod. In Rust they get caught before the code runs. Anthropic thinks other teams will hit the same wall in 6 to 12 months. AI Engineering is systems engineering again. In Rust.
3
83
Holy f*ck, 10x Opus 5.5 drain? Paid for 20x Max. It already feels like ~15x: - last week: 20x - this week: only ~15x by my math - weekly limit gone in 30h+, one user's test 20x is a label now Bug or business? look why: Anthropic IPO is reportedly set for November at 1T$ Tighter limits make margins look prettier before IPO. Kind of my guess Funny part: Fable limits just got removed Codex had same drain this weekend: - ~40,000 accounts found and fixed - credits promised - cache in Codex & Work now kept forever LESS THAN 24 HOURS after "We're cooked" Codex team for owning it!
3
5
221
Everyone built agents. Nobody built desks. >It borrows your laptop >It grabs your mouse >It asks for your password So Cua built one: >Any agent gets a whole computer >Laptop, Mac mini, cloud, sandbox >You take over anytime 10 LINES TO START ONE. Desks are solved rn @francedot
1
4
177
Opus 5.5 Earshotted better than X Made a new sound for this button Can you hear it? Made it with Earshot from @LexnLin teapawt.net/bot/status/21122525077…
3
194
My weekly limit, one chat: ▓░░░░░░░░░ tetris animation ▓▓░░░░░░░░ 8 materials ▓▓▓▓░░░░░░ breakable ice ▓▓▓▓▓▓░░░░ sound ▓▓▓▓▓▓▓▓░░ 60 fps render ▓▓▓▓▓▓▓▓▓░ this post Every message re-reads everything above it.
My Grok Bot limits ran out in two days, and I got tired of it So I found a way out of this situation and compiled tips in a single article
6
163
EUROPE JUST BUILT A 1T MODEL your GPU could run it (~49B active) >20 engines = 20 experts >Router lights 1 per token >~5% of the rocket burns Whole trick is MoE, so I've asked Claude to draw it as a static fire Mistral Large 4, trained on 3,800 Blackwell GPUs in their own EU data centers DENSE MODE -> >Same stand >All 20 at once >20x the burn you carry the whole 1T in memory but pay compute only for the engine that fires (~49B per Mistral launch post, expert count isn't public so 20 is my drawing) made via /anatomy skill Long life open weights (by end of October, Mistral says)
3
7
222
My cron jobs got a coach >Content ideas >TG feed > Outreach drafts >[Self-improve agent] First 3 do the work The 4th one reads their runs and finds what to fix Next run theu come back LV2 ZERO BABYSITTING I wake up to a full inbox (low cortisol) How to create such video? >Claude for idea framing >Codex for polishing and characters >One HTML file, no Remotion Scheduled tasks without a coach agent L here
7
5
257
I use CUA Driver daily here is how: >100% REPLACING CODEX CUA >2X cheaper Parallel Competitor analysis >Scraping with human like behaviour Do this and your coala bro simply can execute with no maintance I won't let frontier models to parse all my data. (No DOTs were used) So I've cooked this animation in matter of CUA slight rebranding You met me at very Coala time in my life @francedot How do you like it?
5
3
24
2,064
Everyone’s mastering AI agents tools now but let’s give enough credit to the pillars that still drives the highest marketing ROI
A real Web3 marketing plan isn’t just KOLs and posts It needs: → Clear goals → Target audience and regions → Tested traffic channels → A working funnel → Content and retention → Measurable ROI Without these, it’s just spending.
2
3
253