Greg Pye

Brand trust will be back for retail IT (and some SaaS). Trust requires legal and reputational cover plus software engineering AND good agentic AI use to defend against threat speed and depth. Vibe coded stuff can't provide it. So, trusted brands, just like for food quality 100 years ago.

What is unclear is who will have earned that trust in the public eye. Apple should. Meta look on the back foot to me, relying on their ad-tech moat for now. Google, probably ok. Anyone ramming AI use rather than brand trust looks vulnerable ... looking at you Adobe.

Brand trust will be back for retail IT (and some SaaS). Trust requires legal and reputational cover plus software engineering AND good agentic AI use to defend against threat speed and depth. Vibe coded stuff can't provide it. So, trusted brands, just like for food quality 100 years ago.

What is unclear is who will have earned that trust in the public eye. Apple should. Meta look on the back foot to me, relying on their ad-tech moat for now. Google, probably ok. Anyone ramming AI use rather than brand trust looks vulnerable ... looking at you Adobe.

If Orbital AI becomes dominant for most inference*, and is not prevented from scaling to 10's of % of AI compute, and AI becomes deeply embedded into orgs, then I would have thought we have a serious risk issue of Kessler events (collision trigger an unstoppable, chain reaction of more collisions).

I would hope for national self interest regulation on this kind of concentration risk, even if the effect is to make this a side-show in capacity terms, meaning terrestrial solutions are needed.

If Orbital AI becomes dominant for most inference*, and is not prevented from scaling to 10's of % of AI compute, and AI becomes deeply embedded into orgs, then I would have thought we have a serious risk issue of Kessler events (collision trigger an unstoppable, chain reaction of more collisions).

UK appears about to have a distracting political context for ages, after which there will likely be more radical change of one sort or another. And Damoclesian high bond prices hover over it all. Why not just cut to a plan to rejoin and help reshape Europe / middle powers?

Bond markets would like it, at least vs. a leadership distraction, and probably in absolute terms as at least 4% of the GDP loss is recoverable (the trade loss part) ... And 4% is HUGE in productivity terms vs. other near term options.

UK appears about to have a distracting political context for ages, after which there will likely be more radical change of one sort or another. And Damoclesian high bond prices hover over it all. Why not just cut to a plan to rejoin and help reshape Europe / middle powers?

Well over half of voters now think Brexit was a mistake, and at 4-8% impact on GDP and negative movement on everything else that was given as a reason, like immigration, it's easy to see why.

Once upon a time Amazon delivery was a step-change better than everyone else (Evri, Yodel, ...). But, my goodness it has degraded. Now, parcels are mostly just chucked behind our garden gate, and flagged as 'handed to resident'. And Amazon customer service don't care as long as it turns up.

@torstenbell.bsky.social said 'life is more complicated than think tank reports', but I am delighted to see him involved directly. That said, the interview on @newsagents.bsky.social felt super partisan (from both sides). Pre MP, I heard balanced thoughtfulness; he is not using his strengths.

AI is not steel or coal or wool or rare earths. It's mass free and moves at the speed. So, why does it matter if we have big #AI data centres in the UK? Data ought not to leak from well designed systems, or at least is no more likely in the USA, Norway (lots of hydro energy), etc. (1/2)

In the UK, defence looks like the straw that breaks the fiscal camel’s back. Lifting to 5% of GDP adds ~£70bn/year. Truss reminded that markets punish overreach fast. So I was wondering - Where would you find £70bn—taxes, pensions, EU growth, spending cuts… or risk a smaller rise?🧵 1/7

Full EU membership may add ~£10bn beyond SM+CU. But would bring euro, rebates, governance questions? We might need it in the end for the £70bn target. Q: Should commit on the big gains now, then let future parliaments decide this, inc. alternative membership options being drafted in the EU now? 6/7

In the UK, defence looks like the straw that breaks the fiscal camel’s back. Lifting to 5% of GDP adds ~£70bn/year. Truss reminded that markets punish overreach fast. So I was wondering - Where would you find £70bn—taxes, pensions, EU growth, spending cuts… or risk a smaller rise?🧵 1/7

Delays to SM/CU entry costs £6–8bn for each year added. We still closely align on regulations, so there is a very credible fast track open. Q: If the gains are known and the costs rise with time, isn’t delay for minor tweaks actually penny wise and pound foolish self-harm? 4/7

In the UK, defence looks like the straw that breaks the fiscal camel’s back. Lifting to 5% of GDP adds ~£70bn/year. Truss reminded that markets punish overreach fast. So I was wondering - Where would you find £70bn—taxes, pensions, EU growth, spending cuts… or risk a smaller rise?🧵 1/7

What about tax and pensions? 2p on basic rate + ending triple lock = ~£32bn (£16bn each). Add £3bn from top rate rise (little headroom as the tax rate is already high). Q: £35bn is only half what is needed, so does this or something similar need to be assumed, even if it does little for growth? 2/7