"It's possible to come up with entirely fictitious solutions to entirely real problems." #digitalsovereign
No way.... #AI #blockchain #crypto #ponzi

If: digital sovereignty + AI = digital sovereignty – AI
Then: AI = 0

"If you want a good "national AI strategy," try this: save your money until the bubble bursts, and then buy your GPUs and hire your talent at 10 cents on the dollar and put them to work refining open source models:"
Thank you so much, @pluralistic !

https://pluralistic.net/2026/06/18/their-trillions-our-billions/

Pluralistic: AI digital sovereignty risk doesn’t exist (18 Jun 2026) – Pluralistic: Daily links from Cory Doctorow

"And in the meantime, please let's stop pretending that digital sovereignty has anything to do with "national AI." If Trump takes away your AI, everything is fine. If Trump takes away your iPhones, Office 365 and tractors, your country grinds to a halt. This is just not that complicated."

@mfeilner i might escalate that to "everyone starves".

You would think the eu of all places would understand the food security implications of how farming equipment now works.

@mfeilner Of course national AI is a digital sovereignty risk. If AI access was shut down today, the fallout would be limited, that's true. Software development productivity would go down in orgs that have adopted AI already, but little more. That is very quickly changing though.

Take just cyber as an example: Today's AI models are already highly capable in the field. More than the median human bug finder. The upcoming rrality of the internet will be that any reachable endpoint will be hammered by AI vulnerability scanners. Controlled by criminals and state actors building initial access databases.

This is already possible today, with most production software having many critical vulnerabilities. It just wasn't economical for threat actors to engage with every possible target. It now is.

The flipside thereof is that security of median applications has to rise dramatically and be kept at that higher baseline level. *That* isn't possible at scale without strong AI.

Then there is also the economic effects of AI in software development. European companies without AI in their processes would have no chance of competing with US or Chinese ones that do. And software development is just the first of very likely many domains where the technology will cause strong productivity increases.

I agree with you on blockchain though. That is a solution in search of a problem. The technology is fascinating in any case.

@jones

"Then there is also the economic effects of AI in software development."

As much as there is no accountable ROI, I do not see any accountable risk. I am with Cory here. Totally.

@mfeilner this METR work does a very good job at researching the impact of AI on software development. The methodology is solid. It is a good indicator that LLM's did indeed reach a turning point in capability in late 2025.

https://metr.org/blog/2026-02-24-uplift-update/#wider-adoption-of-ai-has-made-it-more-difficult-to-measure-task-level-productivity

We are Changing our Developer Productivity Experiment Design

Our second developer productivity study faces selection effects from wider AI adoption, prompting us to redesign our approach.