Mark 🇺🇳

Techie. Built VH1, G-Cloud, Unified Patent Court, UN Global Platform. Saved UK Economy £12Bn+. Building AI stuff #WardleyMaps #PromptEngineer #MLOps #DataOps

[Mark 🇺🇳 @mcraddock]

Here's a Wardley Map on Prompt Engineering, showing the components and skills that are needed [draft].

You can find the source for the Wardley Map here, https://t.co/f9HQviraks #WardleyMaps

[OpenAI @OpenAI]

We’re bringing new capabilities to GPT-Rosalind, a model series purpose-built for life sciences research at enterprise scale.

It brings GPT-5.5’s agentic coding and tool use together with stronger intelligence for drug discovery, analysis, design, and experimental workflows.

https://t.co/SrAJ3Mt7ka

[Mark 🇺🇳 @mcraddock]

So many are thinking of 'Are we ready for AI'

[Mark 🇺🇳 @mcraddock]

Peeps still getting this so wrong.

https://t.co/DrK9uNN67t

[Mark 🇺🇳 @mcraddock]

@emollick Interesting. Obviously don't use any of them. I have Claude Code running solid for 5 hours on a small project that would take a dev weeks ... and that was 6 months ago

[Owen Boswarva @owenboswarva]

University of Exeter is apparently the first public authority to be fined for contempt of court in an FOI case https://t.co/IxCj7Mv6zU (England and Wales)

Cleasby v University of Exeter & Anor [2026] UKUT 187 (AAC) https://t.co/2S3qqBCj1O

FAFO I guess

[Timothy B. Lee @binarybits]

I don't believe any company accidentally spent $500 million on Claude in a month. The number is an order of magnitude too large to be plausible.

[Mark 🇺🇳 @mcraddock]

Why doesn't the World feel any smarter?

[Chris @ChrissGPT]

A million users isn't cool. You know what's cool? A billion .. users

ChatGPT is now being used by 1/8th of the planet per month!! And we’re still in the GPT 5 era.

“roughly three years after launch, surpassing the pace set by apps including Google Maps, TikTok, Instagram and YouTube, Sensor Tower said.”

This does put into perspective the crazy ARR Anthropic has just with 56 million active users per month.

[Mark 🇺🇳 @mcraddock]

@ChrissGPT With a billion people using it, you'd think the World would feel a little smarter 😀

[Mark 🇺🇳 @mcraddock]

@DataChaz Halve life of knowledge https://t.co/vTae9Bo9Nd

[Anthropic @AnthropicAI]

We’re expanding Project Glasswing. We’ve extended access to Claude Mythos Preview to approximately 150 additional organizations, based in more than fifteen countries.

Read more about this expansion and our future plans for Project Glasswing: https://t.co/QrtHSBdRbh

[Chris @ChrissGPT]

Potential Leak: OpenAI might release a site creator tomorrow! Pretty exciting can’t wait to see it in action

This will not be the only thing they’ll showcase ofc

[Mark 🇺🇳 @mcraddock]

@choLord97 @GithubProjects https://t.co/09PCCBErLM

[Mark 🇺🇳 @mcraddock]

@choLord97 @GithubProjects Yes

[Anthropic @AnthropicAI]

Anthropic has confidentially submitted a draft S-1 registration statement to the Securities and Exchange Commission.

Pending completion of SEC review, this gives us the option to pursue an initial public offering.

Read more: https://t.co/onGZAhRLvD

[Mark 🇺🇳 @mcraddock]

@DaveShapi

[Mark 🇺🇳 @mcraddock]

@DaveShapi Mmmm. https://t.co/pemLQpDwAG

[Martin Lewis @MartinSLewis]

Some people question how much impact charities can have... I'm blown away by this summary of @mmhpi's in the 10yrs since it started.

My huge thanks and congrats to the team, trustees, research community & advisory board. Full impact report here https://t.co/cmrQkZ9XVQ

[Carlos E. Perez @IntuitMachine]

Jensen Huang is right. For corporate leadership that is out of ideas, the easiest solution it to reduce the cost of what you are currently doing, rather than exploring opportunities for greater growth. The latter just takes more intelligence (which coincidentally AI provides more of).

[Felix Rieseberg @felixrieseberg]

Also out today: You can now directly configure the effort level and adaptive thinking in Code (/effort) and Cowork!

Effort allows you to tune Claude's intelligence vs token spend, trading off capability for faster speed and lower costs.

Opus 4.8 respects effort levels strictly, especially at the low end. At low and medium, the model scopes its work to what was asked rather than going above and beyond. This is a good option to control spend and cost, but for more complex work we do recommend High and above - your outputs will probably be better.

[Mark 🇺🇳 @mcraddock]

@GithubProjects https://t.co/M2XH4fGTtk