3 ms·

So good to see these developments. Every country should do this. I'd even say every person should gave their own personalized AI running on their own computers.

So good to see these developments. Every country should do this. I'd even say every person should gave their own personalized AI running on their own computers. If only the costs involved were not so astronomical.

nathanielsimard 5mo ago
I think it will be cost effective at some point. Computers were limited to research institutes before the personal computer arrived.
- matheusmoreira 5mo ago
  I hope you're right. I really don't want a future where only corporations and governments have computers.
14u2c 5mo ago
Nvidia will certainly be pleased.
mediaman 5mo ago
Why? That doesn't make any sense. The government would be far better off figuring out how to take commodity models and applying them to government functions where they can, with deterministic scaffolding and guardrails, to make government more efficient, optionally using RL on traces from their use to improve their performance. Imagine taking models and fine-tuning them / doing RL rollouts to help automate permit application approvals, as applied specifically to Dutch permit processes. That would be a real help to Dutch businesses! That type of applied AI is more interesting and effective now than just trying to make another foundational model that isn't going to work well or do anything of economic value.
- matheusmoreira 5mo ago
  > Why? Because then the USA can't just turn it off.
- edg5000 5mo ago
  Another thing they could do is try to attack the the suppply chain issues. Try to form an alliance to block RAM deals or something, or to get fabs on EU soil, making HBM for the people. We have some bargaining chips, especially when banding together with a few large EU states. Not as EU, just a few specic countries. No bureaucracy, just elite trade diplimacy. Probably best done in secret so the big labs don't catch wind of this. Any NL/DE/UK/FR/CH/PL/IT govt people reading this?
jstummbillig 5mo ago
"Champions of a European AI model should ask themselves if a European effort would be more effective than Meta, which this year will spend more on chips ($125 billion) than Germany spends on defense ($114 billion) and offer salaries of over $100 million to attract the best researchers, and is still failing to catch up. Elon Musk tried and failed to build a good AI model." https://www.siliconcontinent.com/p/nineteen-thoughts-on-ai-and-europe?hide_intro_popup=true https://www.siliconcontinent.com/p/nineteen-thoughts-on-ai-a...

nathanielsimard 5mo ago

I think it will be cost effective at some point. Computers were limited to research institutes before the personal computer arrived.

matheusmoreira 5mo ago
I hope you're right. I really don't want a future where only corporations and governments have computers.

14u2c 5mo ago

Nvidia will certainly be pleased.

mediaman 5mo ago

Why? That doesn't make any sense. The government would be far better off figuring out how to take commodity models and applying them to government functions where they can, with deterministic scaffolding and guardrails, to make government more efficient, optionally using RL on traces from their use to improve their performance. Imagine taking models and fine-tuning them / doing RL rollouts to help automate permit application approvals, as applied specifically to Dutch permit processes. That would be a real help to Dutch businesses! That type of applied AI is more interesting and effective now than just trying to make another foundational model that isn't going to work well or do anything of economic value.

matheusmoreira 5mo ago
> Why? Because then the USA can't just turn it off.
edg5000 5mo ago
Another thing they could do is try to attack the the suppply chain issues. Try to form an alliance to block RAM deals or something, or to get fabs on EU soil, making HBM for the people. We have some bargaining chips, especially when banding together with a few large EU states. Not as EU, just a few specic countries. No bureaucracy, just elite trade diplimacy. Probably best done in secret so the big labs don't catch wind of this. Any NL/DE/UK/FR/CH/PL/IT govt people reading this?

jstummbillig 5mo ago

"Champions of a European AI model should ask themselves if a European effort would be more effective than Meta, which this year will spend more on chips ($125 billion) than Germany spends on defense ($114 billion) and offer salaries of over $100 million to attract the best researchers, and is still failing to catch up. Elon Musk tried and failed to build a good AI model." https://www.siliconcontinent.com/p/nineteen-thoughts-on-ai-and-europe?hide_intro_popup=true https://www.siliconcontinent.com/p/nineteen-thoughts-on-ai-a...