No fresh buckets — last snapshot is old (ingest gap, not silence).
Partial day — headline is the last complete day (2026-10-06).
twitter
tracked since 2026-10-03T04:31:44.334755+00:00
conclusion
Measured mentions per day
day (UTC)
mentions
2026-09-27
0
2026-09-28
0
2026-09-29
0
2026-09-30
1
2026-10-01
6
2026-10-02
29
2026-10-03
36
2026-10-04
41
2026-10-05
42
2026-10-06
2
derived from
letlet connect
word-boundary match over title + text, query='timeline' · window
2026-09-27 → 2026-10-07
· 224 matching posts
evidenceShowing 20 newest of 224
•
twitterbitsbyritik1wnegative
Target to hit 600 followers today
Looking to #connect with:
- Builders 🏗️
- Developers 💻
- Founders 🤬
- Vibe coders 😎
- AI nerds 🤓
- Indie Hackers 🪂
- Designers 🖼️
If you're building something cool, please do share what you are building & let's connect https://t.co/Xo2bckokjcTarget to hit 600 followers today
Looking to #connect with:
- Builders 🏗️
- Developers 💻
- Founders 🤬
- Vibe coders 😎
- AI nerds 🤓
- Indie Hackers 🪂
- Designers 🖼️
If you're building something cool, please do share what you are building & let's connect https://t.co/Xo2bckokjc
Key takeaways for marketers:
1. If your app is coming back, say what's back first. People want to know what works again before they hear about anything new
2. Build each post around one real moment. A surprise talent, a fail, a funny reaction... don't explain the app, instead, let people see it working in the backgro…Key takeaways for marketers:
1. If your app is coming back, say what's back first. People want to know what works again before they hear about anything new
2. Build each post around one real moment. A surprise talent, a fail, a funny reaction... don't explain the app, instead, let people see it working in the background
3. Use groups. A social app makes sense when people see several friends using it together
4. Give each type of post one job:
- surprising moments get attention
- friends reacting together prove the app works
- one clear ask gets the download
- regular events keep ppl coming back
5. Don't push for growth before the product is ready. If the app isn't fully working yet, use your account to listen. Read comments, see what people ask, and post when you have something they can use
One of the things I picked up from scaling billing at Flexprice is a simple way to think about pricing architecture:
Start with your core monetization motion.
Let's say it is usage-based.
You build a pricing layer around it - overages, minimums, commitments - all anchored to that same motion.
The invoice tells a st…One of the things I picked up from scaling billing at Flexprice is a simple way to think about pricing architecture:
Start with your core monetization motion.
Let's say it is usage-based.
You build a pricing layer around it - overages, minimums, commitments - all anchored to that same motion.
The invoice tells a story and your contract is a pricing document.
Using these together through the billing stack creates an almost invisible but present revenue logic that compounds over time.
It works especially well for SaaS products where discounts and custom deals can easily dominate, making it hard to maintain pricing integrity without things breaking downstream.
Claude Code is not supposed to make films.
I let it run anyway for a day.
It made a 10-minute Bruce McLaren documentary, with real archive footage and his own voice.
The film, the workflow, and all the tools below 👇 https://t.co/HhUsZkZfZdClaude Code is not supposed to make films.
I let it run anyway for a day.
It made a 10-minute Bruce McLaren documentary, with real archive footage and his own voice.
The film, the workflow, and all the tools below 👇 https://t.co/HhUsZkZfZd
This is exactly the right direction.
Auto-review being free is a huge win. Let agents actually work for hours without constantly asking me for permission, while a second agent quietly catches the genuinely risky stuff.
More autonomy, less babysitting, without throwing safety out the window.
This is how agents become…This is exactly the right direction.
Auto-review being free is a huge win. Let agents actually work for hours without constantly asking me for permission, while a second agent quietly catches the genuinely risky stuff.
More autonomy, less babysitting, without throwing safety out the window.
This is how agents become actually useful.
The first thing I let Opus do is fire this person instantly. Unbelievable soydevism from the highest tier.The first thing I let Opus do is fire this person instantly. Unbelievable soydevism from the highest tier.
I'm 29!
Solo founder from Nepal 🇳🇵
Just launched PDF Kit Pro to make editing scanned documents much easier. Only 5$ for first 50.
>https://t.co/d9XNPEIn5P
Looking to connect with more people in:
• Finance 📊
• Accounting
• Chartered Accountancy (CA)
• Tax & Compliance
• FinTech
If you're in finance or a CA, let's co…I'm 29!
Solo founder from Nepal 🇳🇵
Just launched PDF Kit Pro to make editing scanned documents much easier. Only 5$ for first 50.
>https://t.co/d9XNPEIn5P
Looking to connect with more people in:
• Finance 📊
• Accounting
• Chartered Accountancy (CA)
• Tax & Compliance
• FinTech
If you're in finance or a CA, let's connect! 🤝
Marketing & distribution > just building. 🧠
Studying how brand image is built on Reddit, Insta & YouTube...the platforms, the psychology, what moves the needle.
Founders who win obsess over distribution too.
Let's gooo #buildinpublic #distribution #marketing #reddit #insta https://t.co/yZhhVK5cBiMarketing & distribution > just building. 🧠
Studying how brand image is built on Reddit, Insta & YouTube...the platforms, the psychology, what moves the needle.
Founders who win obsess over distribution too.
Let's gooo #buildinpublic #distribution #marketing #reddit #insta https://t.co/yZhhVK5cBi
Hermes Wingtips #93: let your Hermes Agent install its own skills
your Hermes Agent can find a skill and run the install itself, so you don't need the skill's name or the command. ask in plain words:
"find a skill for flashcards and install it."
it picks a match, runs the security scan and then tells you what to say…Hermes Wingtips #93: let your Hermes Agent install its own skills
your Hermes Agent can find a skill and run the install itself, so you don't need the skill's name or the command. ask in plain words:
"find a skill for flashcards and install it."
it picks a match, runs the security scan and then tells you what to say to use the new skill.
handy when you want your agent to pick up something new, or when you're setting up a fresh profile!
"I bet your agent can't post to Instagram?"
mmmm, let me tryyy👀💅 https://t.co/PuRR4AtnTq"I bet your agent can't post to Instagram?"
mmmm, let me tryyy👀💅 https://t.co/PuRR4AtnTq
comment your website, and I will send you 15 hot prospects ready to buy and showing buying signals
no personal websites, please
let's gocomment your website, and I will send you 15 hot prospects ready to buy and showing buying signals
no personal websites, please
let's go
Another beautiful day to remind you that I'm a visual and brand identity designer ❤️.
I create identities and visual systems for business, startups and brands to help them stand out and attract their target audience.
Let's work on your next project ☺️. https://t.co/w48S3KfxTUAnother beautiful day to remind you that I'm a visual and brand identity designer ❤️.
I create identities and visual systems for business, startups and brands to help them stand out and attract their target audience.
Let's work on your next project ☺️. https://t.co/w48S3KfxTU
It's time to stop using slow programming languages.
I don't think you understand the magnitude of this situation.
Let me explain. https://t.co/6eb2M99jMSIt's time to stop using slow programming languages.
I don't think you understand the magnitude of this situation.
Let me explain. https://t.co/6eb2M99jMS
You think you're running 1000 ads but Meta ONLY sees 100
Meta groups visually similar creatives under one Entity ID, even when they're separate Creative IDs in your account
Everything under one Entity ID shares its learnings, delivery, and costs
So 30 statics of the same product at different angles count as 1
Ads M…You think you're running 1000 ads but Meta ONLY sees 100
Meta groups visually similar creatives under one Entity ID, even when they're separate Creative IDs in your account
Everything under one Entity ID shares its learnings, delivery, and costs
So 30 statics of the same product at different angles count as 1
Ads Manager shows you rows, and rows hide how alike your ads look
So we built Odylic Constellation, a free Mac app that puts every Meta ad you're running into one 3D view
- See every live creative side by side instead of scrolling a table
- Ads running the same visual stack into one card with a count, so repeats are obvious
- Regroup by format, angle, persona, creator, or hook, detected from your ad names
- Every ad sits in the funnel by where Meta spent on it, new people, engaged non-buyers, or existing customers
- Click any ad for daily spend, ROAS, segment split, placements, and video retention
Runs locally, read-only, and installs with one line in Terminal
Let me know if you want a copy
pro tip: do not let AI take away the serendipity in your life. make a concious effort to stumble upon something interesting every single day.pro tip: do not let AI take away the serendipity in your life. make a concious effort to stumble upon something interesting every single day.
Builder Arc - Learning's sharing time (Day 2)
I launched lafda dot fun on Oct 4 at 12:45 PM IST, and it's been roughly 43 hours.
> 20.2k pageviews,
> 7.1k sessions,
> 6.1k unique visitors,
> average session duration 2 min 40 sec,
> 94% of traffic from India, 64% on mobile,
> 120 sign-ups,
> 11 Lafda submissions,
> 0 …Builder Arc - Learning's sharing time (Day 2)
I launched lafda dot fun on Oct 4 at 12:45 PM IST, and it's been roughly 43 hours.
> 20.2k pageviews,
> 7.1k sessions,
> 6.1k unique visitors,
> average session duration 2 min 40 sec,
> 94% of traffic from India, 64% on mobile,
> 120 sign-ups,
> 11 Lafda submissions,
> 0 premium users so far (excluding me, of course).
Now let's talk about bugs and glitches.
I shipped 8 PRs since the last post, and 4 of them were fixes.
From logging and observability, I caught:
> One backend error was 97% of all my errors. Signed-out readers on a post page hit an account endpoint, got "not signed in" 3,867 times, and some saw a "Sign in required" toast without clicking anything. Fixed. It dropped from 3,738 a day to 2.
> Every backend error was logged twice. Now it's logged once. (i forgot to turn of STDIO it in prod)
> Failed sign-ins left no trace in my logs. Now they do.
> Hydration error on the post page for signed-in readers (the page rendered twice). Fixed.
> I removed my auth rate limit. Behind Cloudflare, many users shared one IP bucket, so real users could get blocked. Instead moved rate limit to WAF.
> Big pastes failed to save (over 400 edits at once). Fixed. Only 1 user hit it (me ofc)
From user reporting:
> None
Also shipped: notifications for comments and upvotes, a better paywall, an About page, and writing guidelines & enabled web vitals.
I will report back with the analytics, issues, and monetization details.
Let me show you something crazy.
This is Qwen 3.8 27B, and every frame is 1 real equation from the model's actual math. The prompt is easy and thinking is off: we ask what the capital of France is, with a one-word reply.
If you think this video is doing math fast, it's actually skipping about 7.2 MILLION equations be…Let me show you something crazy.
This is Qwen 3.8 27B, and every frame is 1 real equation from the model's actual math. The prompt is easy and thinking is off: we ask what the capital of France is, with a one-word reply.
If you think this video is doing math fast, it's actually skipping about 7.2 MILLION equations between every frame. After all the math, Qwen gives the final answer: "Paris".
That one-word answer took 25.8 BILLION equations, and my NVIDIA RTX PRO 6000 Blackwell did it in about 13 milliseconds. This video is 32 billion times slower than the actual GPU. Showing every single equation at one per frame at 60 fps would take over 13 years, for one word.
Now imagine scaling that from 27B parameters up to models with hundreds of billions or even trillions of parameters, and that's why AI needs so much compute.
Appreciate your GPUs, no matter which one you have. They're actually insane.
RT @matteing: We weren't kidding. Let's talk about Space. ✨
Have you tried Space yet?
What's good or bad about it?
What could we improve?…RT @matteing: We weren't kidding. Let's talk about Space. ✨
Have you tried Space yet?
What's good or bad about it?
What could we improve?…
Hey uh, @Keurig - what the hell is a COFFEE MACHINE uploading, let me double check...
ONE FUCKING TERABYTE OF DATA IN 10 DAYS!?!?
Immediately unplugging that. Getting my parents a new coffee machine. https://t.co/tTHMYC7hNDHey uh, @Keurig - what the hell is a COFFEE MACHINE uploading, let me double check...
ONE FUCKING TERABYTE OF DATA IN 10 DAYS!?!?
Immediately unplugging that. Getting my parents a new coffee machine. https://t.co/tTHMYC7hND