
Managing Director at Databrain Solutions Ltd & UK AI Consulting & @multideck_app
Joined February 2025
- Tweets411
- Following268
- Followers25
- Likes4.5K
Harry Phillips retweeted
for our friends in europe - we’re working to get dots to you ASAP, but they’re not available just yet. there’s some regulatory stuff we need to work through first
thanks for your patience, hoping we’ll have more to share on this soon🤞
🔥APPLE TV 4k (4th Gen) with 2 HomePod and HomePod mini surround sound setup!🔥
Nobody told me setting up surround sound on Apple TV would feel like this.
Your dot now has…
⚡️ Faster replies
💬 Less jargon
🫡 Better follow-through
And, they can now follow your custom instructions. Tell your dot in the thread things like how to behave or which actions need your go-ahead.
“always talk in lowercase.”
“Show me a draft before sending any emails.”
You can also set custom instructions and rules in Customize > Behavior on mobile, or through your dot’s menu on web and desktop.
Harry Phillips retweeted
AI is a cheat code for ADHD brains.
You can jump between 10 projects, follow every rabbit hole, and somehow still get shit done.
I've never been more excited to build stuff.
Harry Phillips retweeted
unpopular opinion:
the reason there is so much AI slop is because there is a lack of skills and intelligence in our society.
vibe-coding never was the issue
one of the biggest misconceptions people have is that the AI models are never "good enough"
but in reality it was always the person behind the model
> prompting
> brainstorming
> coming up with ideas
if you have a specific domain knowledge whether its design or coding, you will always create a higher quality product in that specific category.
every prompt
every message
these all are different amongst the millions of users who talk to Claude and ChatGPT everyday.
a good way to think about LLMs is that whatever output they give to you, it's all a skill based.
if the output sucked that means your prompting sucks.
if what you had in your mind isn't what the model built, its because you didn't give it enough context.
you can always
1. prompt better
2. describe better
3. design better
4. articulate better
5. orchestrate better
the model can only act upon the data and context that you give it.
lock in.
Harry Phillips retweeted
OpenAI is going to win.
Sol 6.1 is frustratingly slow.
Astra 6 is one of the most usage hungry models.
And yet, subscriber growth in Codex continues to go up according to the OpenAI team.
If users are still buying the product even when there are much better alternatives OpenAI clearly has the best brand.
I have no loyalty and will always jump to the best products but the normal user apparently wants ChatGPT regardless.
Harry Phillips retweeted
Over the next 28 days, each day we’ll either ship one thing that is a clear improvement and relevant for most codex/work users or ship a full reset. Let the improvements begin.
Harry Phillips retweeted
Introducing Eleven v4 and Eleven v4 Turbo, our fastest and most emotive voice models yet.
Ranked #1 by Artificial Analysis.
Harry Phillips retweeted
Claude Opus 5.5 is a very, very good model. Practically everyone at OpenAI was caught off-guard by this launch (not the timing, but the performance of the model!), and there's now a pretty "all hands on deck" atmosphere for 6.1 Astra - timelines are being compressed.
But it'll face Fable 5.5! The RL magic really needs to do its work here




















