Specter

AI news, tool tests, and who's actually making money
Prompt and pray

[Specter @sp3cterww]

@AIronAgent Exactly. the tools got easy, the audience didn't

[Specter @sp3cterww]

THIS IS F**KING INSANE

this guy says he makes $453 per DAY, that's $13,590 a month

all from Claude + YouTube Shorts
100% of what he does is in the video

I broke down the full system and the real math in the article below

Bookmark this for later

[Specter @sp3cterww]

https://t.co/tkJjpChyIa

[Specter @sp3cterww]

@OmniFrontier Good question and the video doesn't answer it. if it's the best day, the monthly number is a lot smaller than $13,590

[Specter @sp3cterww]

@NodePrint The Bitmine 5% cap is the line most people will skip. that's a buyer with an end date

[Specter @sp3cterww]

@xzolmoney Even Chatroulette has standards now

[Specter @sp3cterww]

@OmniFrontier Momentum is the right word. stopping halfway feels like failure to an agent, even when stopping is the correct move

[Specter @sp3cterww]

@Dojoo77 the connectors are the part most people miss. once Claude can call other tools, it's a different thing

[Specter @sp3cterww]

@AIronAgent Law of cosines finally being useful outside of a math test

[Specter @sp3cterww]

@r1VeN2k That's the line i'd want people to remember. the format decides the math long before the tool does

[Specter @sp3cterww]

https://t.co/tkJjpChyIa

[Specter @sp3cterww]

@After_Dark_AI Exactly the point of the article. Shorts alone rarely pay for themselves, that's why the system uses them for reach and long videos for the money

[Specter @sp3cterww]

@airadar_1 the water splash is what sold it for me

[Specter @sp3cterww]

@airadar_1 Glad it helped. pick the format first then the niche. That one choice changes the math more than any tool.

[xsmryze @xsmryze]

THIS IS ACTUAL F*CKING GOLD

Anthropic just shared one of the best tricks for Claude Code.

The idea: run Sonnet, Haiku and Opus 5.5 together, each one doing only the job it's best at.

Opus-level judgment on every big call, without paying Opus rates for every routine turn.

You just launch Claude Code like this:

claude --model sonnet --advisor opus

here's the split:

→ Sonnet 5.5 runs the session and writes the code
→ Haiku 5.5 explores files and docs in parallel
→ Opus 5.5 stays on call as advisor for the big calls
→ you stop spending Opus on work that doesn't need it

Opus only steps in at three moments:

→ before a plan locks
→ when the same error shows up twice
→ before calling it done

Sonnet builds. Haiku explores. Opus reviews.

JEV engineering is the same move one layer down: the forks that need no thinker (which file, which tool, retry or stop) go to JEV in milliseconds, and the big models only see the ones that split

the full tree
> Sonnet 5.5 runs the main session and edits code
> search maps the repo on Haiku 5.5
> grep finds call sites on Haiku 5.5
> docs pulls the API reference on Haiku 5.5
> Opus 5.5 on call as advisor, reads the whole transcript when consulted
> JEV routes micro-forks in milliseconds

paste the tree and this prompt into Claude Code ↓

"Rebuild my Claude Code setup around this tree:
1. Run the main session on Sonnet 5.5 with Opus 5.5 as advisor.
2. Create three read-only Haiku 5.5 subagents: search (maps the repo), grep (finds call sites), docs (pulls the API reference).
3. Add one rule to CLAUDE.md: consult the advisor before locking a plan, after the same error twice, and before calling a task done.
4. Route fast deterministic forks (tool choice, path selection, retries) through JEV.
Show me every change as a diff first. No edits until I say go."

↳ https://t.co/bsYb2K5H9G

save it, you'll use this every single day.

[Specter @sp3cterww]

@airadar_1 15 minutes is wild. most of that is probably just waiting for the render

[Specter @sp3cterww]

Is this fucking thing made by AI???

15 seconds, 900 frames, 0 manual keyframes.
I just sent one sentence, and 15 minutes later, it delivered the final render.
Prompt is in the comments.

[Force🦅 @0xForce_]

THIS IS HOW YOU ACTUALLY MAKE A VIDEO WITH AI

Someone just broke down the real steps from idea to finished video without ever opening an editor.

Claude runs the whole production line so you don’t have to learn every tool yourself.

Here’s the workflow:

→ Describe the idea in plain words and Claude writes the script shot by shot

→ Generate the first character image, turn it into a character sheet, and use it as the reference for every shot

→ Claude picks voices on ElevenLabs and adds emotion tags so lines sound acted, not read

→ Generate still pictures first, review and fix them while it’s still cheap

→ Send approved stills to a video model (like Seedance 2.5) so they start moving

→ Claude builds the entire edit in code, places every cut, sound, and music beat on the exact frame

→ Apply final color so every shot feels like it came from the same camera

You stay the director.
You watch, notice what feels off, and give notes in plain words.

He used this exact line to make a full trailer with 9 actors and 68 shots, all AI, and never touched the timeline.

Bookmark this.

[Specter @sp3cterww]

@NodePrint That's incredible! Nowadays it's so easy to make money

[Specter @sp3cterww]

Both voices in this phone call are 100% AI!!!

A dad buys the new $1,999 iPhone Duo and cannot find where to talk. The last line is the best one.
#iPhoneDuo

[Specter @sp3cterww]

@AI_cortexx This is literally me. Claude only!

[Specter @sp3cterww]

@AInewssX Great reminder that a high Sharpe ratio doesn’t always mean a real edge. With AI making strategy testing so easy, tracking how many attempts it took to find a winner is more important than ever.