We sold DeepMind to America for four hundred million pounds. Now a politician is selling our NHS data. We have done this before, and we never seem to learn.
Your argument that the Deepmind sale was for too little or shouldn't have happened at all assumes that Deepmind could have been a success on its own (remember that it had no products and no revenues when it was sold) and neglects the billions Alphabet has poured into Deepmind and the UK AI ecosystem since the acquisition. Deepmind is a jewel *because* Alphabet had supported it so fullsomely, not despite that.
I think that proves my point rather than defeats it.
Of course DeepMind needed huge capital, compute and long-term backing. Frontier AI was never going to scale like a normal software business. The fact it had no conventional product or revenue in 2014 is precisely why patient capital mattered.
Alphabet backed it brilliantly. That is the indictment of Britain.
Why was that scale of ambition available from a US hyperscaler, but not from British institutions sitting on vast pools of capital?
Google reportedly bought DeepMind for around £400m. If DeepMind were independent today, given AlphaGo, AlphaFold, Gemini, Nobel Prize-winning science and the current valuations of frontier AI labs, it is not absurd to posit it as a $1tn strategic asset — roughly a quarter of UK GDP.
Jobs staying in London is good. But jobs are not the same as ownership of a trillion-dollar capability.
Britain retained activity. Alphabet captured the ownership, IP, roadmap, compute integration, platform leverage and balance-sheet upside.
So no, I am not saying Google added no value. Clearly it did. I am saying Britain created the asset, failed to back it at scale, and then watched the real upside compound somewhere else.
That is exactly the mistake we should not repeat with NHS data.
Your argument that the Deepmind sale was for too little or shouldn't have happened at all assumes that Deepmind could have been a success on its own (remember that it had no products and no revenues when it was sold) and neglects the billions Alphabet has poured into Deepmind and the UK AI ecosystem since the acquisition. Deepmind is a jewel *because* Alphabet had supported it so fullsomely, not despite that.
I think that proves my point rather than defeats it.
Of course DeepMind needed huge capital, compute and long-term backing. Frontier AI was never going to scale like a normal software business. The fact it had no conventional product or revenue in 2014 is precisely why patient capital mattered.
Alphabet backed it brilliantly. That is the indictment of Britain.
Why was that scale of ambition available from a US hyperscaler, but not from British institutions sitting on vast pools of capital?
Google reportedly bought DeepMind for around £400m. If DeepMind were independent today, given AlphaGo, AlphaFold, Gemini, Nobel Prize-winning science and the current valuations of frontier AI labs, it is not absurd to posit it as a $1tn strategic asset — roughly a quarter of UK GDP.
Jobs staying in London is good. But jobs are not the same as ownership of a trillion-dollar capability.
Britain retained activity. Alphabet captured the ownership, IP, roadmap, compute integration, platform leverage and balance-sheet upside.
So no, I am not saying Google added no value. Clearly it did. I am saying Britain created the asset, failed to back it at scale, and then watched the real upside compound somewhere else.
That is exactly the mistake we should not repeat with NHS data.