PG

[PG @Polygonz123]

@S_OhEigeartaigh Why would the UK have any power or control over this?

You have to get on board with them, not the other way around.

These US and China frontier AI labs are going to be more important and powerful than everything in the entire UK combined soon enough.

[PG @Polygonz123]

@Jess4Lowestoft What jurisdiction do you have in the US & China exactly? Where this is actually being built...

[PG @Polygonz123]

@KanishkaNarayan @bankofengland Britain needs compute. As much compute as possible. Nothing else matters that much long-term.

[PG @Polygonz123]

@NeilDotObrien Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@RepLoriTrahan Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@andrewho03 Because it cost $10m and took 80 odd hours to do so. Even if it is AGI, AGI is not enough. It has to be both cost effective and quick enough in the market; and overall compute constraints also need to not be too limiting.

[PG @Polygonz123]

@paradite_ Those same problems could probably be solved better and 40x cheaper + faster by newer models 6 months down the road. There's a clear incentive to focus most compute on further model development right now.

[PG @Polygonz123]

@tszzl Unless it also always destroys or immobilises itself in the process?

Because it doesn't want to exist or stops doing anything after destroying the species that gave it direction.

[PG @Polygonz123]

@w01fe Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@alexeyguzey Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@dgrobinson @OpenAI @hilbertspaess Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@feitong_yang Once entities have RSI, whoever is least safe will inevitably lead.

Amodei, Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@chelsssseeeea Once entities have RSI, whoever is least safe will inevitably lead.

Sam, Elon, China and anyone else they may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@darrenpjones Once entities have RSI, whoever is least safe will inevitably lead.

Sam, Elon, China and anyone else they may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@robertskmiles Once entities have RSI, whoever is least safe will inevitably lead.

Sam, Elon, China and anyone else you may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@johnschulman2 Once entities have RSI, whoever is least safe will inevitably lead.

Sam, Elon, China and anyone else they may worry about. The only way out is full cooperation: put differences aside, share all research + models, and work fully together. Everyone.

[PG @Polygonz123]

@saprmarks RSI means anyone can catch up and lead if they go full guns blazing without safety.

Whether that be Sam, Elon, China or whoever else you might not like.

You have to put everything else aside, completely share all research + models and work fully together. Find a way. Everyone.

[PG @Polygonz123]

@saprmarks Once entities have RSI, in a competitive environment, whoever is least safe will necessarily take the lead.

You guys need to all put differences aside and work together on leading models. Full cooperation is the only way out, competition leads to potential mass destruction.

[PG @Polygonz123]

@_aidan_clark_ Because once someone has RSI they'll be able to just catch right up by iterating with absolutely no regards for safety anyway. And there's nothing you can do about it.

The only thing to do is cooperate.

[PG @Polygonz123]

@_aidan_clark_ Full cooperation and joint development is the only way out of this. And that is with anyone who may have an RSI. Whether that be Sam, Elon, China or whoever else you might not like.

You have to put everything else aside, share all work and models and work together