Solo builder: https://xitter.gq/t.co/cnNroE50wN + https://xitter.gq/t.co/bDu4YLg6ae + https://xitter.gq/t.co/BVwkRvCR0w
Rockingham, Perth (WA)
Joined February 2008
- Tweets9K
- Following1.3K
- Followers1.7K
- Likes5.1K
Pinned Tweet
Every CEO on the planet regardless of industry or company size, needs to do these 3 things:
- Spend $20 to get ChatGPT Plus and experiment with the Advanced Voice feature.
- Book a demo of the Apple Vision Pro at their nearest Apple Store.
- Catch up on the last 12 month of robotic advancement. OpenAI Figure and Tesla Bot are a good place to start.
That kinda of visualisation UI reminds me of the Winamp's skins (skins.webamp.org/). Completely useless but somewhat desirable :D
Winamp Skin Museum
Infinite scroll through 100k Winamp skins with interactive preview
skins.webamp.org
holy sh*t. Official Anthropic tip for Claude Code: stop burning Opus 5.5 context on tasks Haiku 5.5 can swarm, while Sonnet 5.5 builds
put the architect on call with /advisor
claude --advisor opus --subagents haiku
Sonnet 5.5 drives the primary session on high effort, writing diffs and running test suites
Haiku 5.5 subagents swarm the repo in parallel at 340 tok/s, handling file discovery and spec docs lookup
Opus 5.5 stays on call in the background as the advisor, only stepping in at three critical moments:
→ before a plan locks: does the implementation plan miss auth invariants or schema contracts?
→ when a test breaks twice: are we fixing the root cause or falling into a recursive rabbit hole?
→ before calling done: did the full diff introduce hidden regressions or break pre-flight rules?
Opus 5.5 reviews. Sonnet 5.5 builds. Haiku 5.5 swarms
Jev engineering runs the identical pattern one layer down: mechanical decisions that need no reasoning (which file to open, which tool to invoke, retry or abort) execute in 16ms, so the frontier models only wake up when execution paths actually diverge
Plan on high. Delegate on medium. Keep Opus on call.
- the full advisor setup
> Opus 5.5 on call reads full session history and catches deep architectural traps
> Sonnet 5.5 lead drives edits, writes core logic, and executes test harnesses
> Haiku 5.5 subagents swarm AST parsing, grep scans, and API docs in parallel
> JEV micro-fork layer resolves 1,500+ mechanical routing branches in under 16ms
> advisor stays completely silent on routine bash execution to protect context
paste the setup and this prompt into Claude Code below:
"Configure my Claude Code workspace for hierarchical advisor orchestration:
1. Audit ~/.claude/settings.json and project config for model roles fitting lead, subagents, and advisor:
> Pin main session model: sonnet with effortLevel: high for primary execution
> Pin subagents model: haiku with effortLevel: medium for parallel file discovery and docs retrieval
> Pin advisor model: opus on call for strategic review
2. Enable the advisor tool in ~/.claude/settings.json:
> Set advisorModel to claude-opus-5-5
> Enable subagent parallel dispatcher pool (3x workers)
3. Configure automatic advisor consultation checkpoints in ~/.claude/CLAUDE.md:
> Consult /advisor opus before finalizing multi-file architecture plans
> Automatically summon /advisor opus when the same test or compiler error fails twice
> Enforce advisor diff contract audit before declaring tasks complete or staging git commits
4. Constrain subagent scopes:
> Haiku subagents return structured AST summaries and docs snippets only without modifying main session context
Show every configuration diff first. Do not apply edits until confirmed"
↳
code.claude.com/docs/en/advi…
@Scobleizer I don’t understand how/why this👇 is not a thing in VR yet? Seems to me like an obvious banger for Apple (and others) yet no signs of it.
On the topic of use case for the Apple Vision Pro, since nobody asked for it here are my 2 cents... I don’t want to work on them, I think it won’t add to my working comfort however I’m excited about watching movies on them ski goggles. I think it will kill cinemas. Why pay $20 to go to a movie theatre when you can pay less and be actually physically immersed in the movie in the comfort of your home.
Purposefully shot movies will blow our minds, providing true immersive experience where the storyline and photography actually build on the immersion. This will bring blockbusters to a whole new level (think the Avatars, Marvels, Transformers etc.), could also be awesome for all sorts of shows (theatre, music concerts…). Here is the kicker though, if you're worried about the format and lack of movies "purposefully shot for the AVP" consider that we are merely month away from LLM (<cough> AI </cough>) being able to generate the additional visual data to bring key movie elements to your environment. For example: Imagine watching Aliens (1986) and your whole living room gets covered in organic alien structures, it’s dripping and disgusting and it closes up on you as the motion tracker starts to beep… How about watching Titanic and you’re actually seating inside the ship? Gladiator and the Colosseum crowd is brought into your living room, literally seating on the couch with you as Commodus is getting the pointy end of the spear… And one of my favourite would be to watch the movies with the cast / director seating next to you and providing you with comments and anecdotes… How awesome is that going to be?
Has Apple quietly reinvented how we consume visual entertainment? I think it might have. If they ask for a 30% clip (and yes they will, justifying it by the fact that data is generated on the fly by the device) it’s going to suck for Netflix, Disney, Amazon Prime etc. until they release their own googles? maybe?
I don’t see how cinemas would compete? Might suck a little also for TV manufacturers trying to pump us extra inches of OLED HD something.
There you go, that was my 2 cents that nobody asked for :)
I find it amusing that suddenly we all seem to care so much about code quality and dread the “ai slop”.
Mate… most (if not all) of the popular apps we use daily have always been riddled with bugs (looking at you Meta) and run on spaghetti code from decades of language and infrastructure changes.
Say what you want about Anthropic but one good thing that came out of their computing shortcomings is that Claude is vastly more token efficient than Codex. Codex is an absolute token ogre in comparison, easily 10-20x less efficient than Claude.
I suspect that if Claude was to be let loose it would outperform Codex easily by some margin.
I’m tired of my x feed essentially turning into bloody tik tok… @nikitabier could we have a slider to choose what percentage of posts in our feed should show a video?
I think @EmmanuelMacron read your posts @gearside ;)
Les choses avancent, mais pas assez vite. Jamais assez vite.
J’ai donc pris un engagement clair : accélérer, une bonne fois pour toutes, la réalisation de nos projets industriels sur l’ensemble du territoire.
Nous avons montré au monde que nous étions capables de rebâtir Notre-Dame en cinq ans. C’est cet esprit que je veux retrouver aujourd’hui.
Aujourd’hui, nous identifions 150 grands projets, 150 cathédrales industrielles, partout en France que nous allons faire avancer beaucoup plus rapidement.
Cela représente 71 milliards d’euros d’investissement et plus de 32 000 emplois, au service de notre souveraineté.
Et cela s’accompagne de très nombreuses simplifications pour faire mieux et plus vite.
Nous accélérons.
I found a new way to benchmark LLMs: get them to work out the cost of parking at the Cannes Palais des festivals car park.
Prompt: Can you work out parking cost on Friday from 10am to 3.30pm.
Claude Opus 4.7 adaptive: 9 euros
ChatGPT Thinking Extended: 8.20 euros
Grok: Fail, "High demand, wait or pay".
I’m currently in France visiting family and the AI / LLM adoption in businesses is hugely slowed by data sovereignty issues. Businesses cant use SOTA models because they are not hosted/hostable in France.
For anyone under 50 (ish) tomorrow is set (🤞) to be the biggest human accomplishment that we’d have witnessed in our lifetime. Go Artemis Go you good thing!!!
So now that Twitter/X is more expensive than a Claude Code subscription what is (if any) the alternative for X Pro? I want alerts on certain terms for example.
What’s a better fit to work with LLMs:
0%Technical product mnger
100%Product minded SWE
2 votes • Final results