This week a major US AI development flew under the radar - a US congressional commission, the U.S.-China Economic and Security Review Commission (USCC) presented their annual report to congress.
Their number one recommendation was that Congress establish a "Manhattan Project-like program dedicated to racing to and acquiring an Artificial General Intelligence (AGI) capability".
There's a lot to say about this, which I cover in the piece.
It's an awfully big logical stretch from "a report used the phrase Manhattan Project" to "AGI may be born secret." This is coming from someone who knew the relevant history/concept beforehand.
The word "may" is doing a great deal of work in your headline.
It's a good idea. I don't think secrecy is really necessitated aside from the idea that if the US government is the one spending the money, then they should reap the benefits. Ideally, I'd love to see many governments collaborate here and share the technology with each other as a result.
What is scary about this memo, and contrary to your "lets be friends" outlook, is that we are entering a geopolitical era of precisely the opposite.
The US is withdrawing from its global cop role. Russia and China are saber rattling. NATO may be dissolved by Trump, at least the American side of it. Israel and Saudi Arabia are new strange bedfellows and allies. Japan and India are vastly increasing their military spending, and Japan will likely arm itself with nuclear weapons. Sunnis and Shiites are nearing a war. Turkey wants to throw its weight around regionally. International trade is very likely to be strongly disrupted and production nearshored.
On this point - check out Matthew Pines on his view that over the next year the US will force allies to convert treasuries and gold to US century bonds, with the stick being a threat of being left outside the US security and technology alliance system. The conversion being a way to deal with the big overhang of US short-dated debt.
I don’t care who is researching this, we won’t have AGI by 2027 and super intelligence by 2030s.
> Based on trends in AI capabilities research since GPT-2, we are on course to expect AGI by 2027. Once AGI capability is available, if labs focus on automating AI research itself, progress in AI should accelerate. If similar progress can be achieved as the phase from GPT-2 to GPT-4, or GPT-4 to AGI, we should expect Superintelligence before the end of the decade.
"born secret", maybe. But by a US government program? Lately besides corruption and inefficiencies, the government didn't have much to show. NASA go trounced hard by SpaceX.
The Kamala "broadband" project did cost $42 billion to taxpayers and not one home has been connected with this new program. In nearly four years. Zero. Meanwhile StarLink is actually connecting people.
Be it NVidia, OpenAI, Meta, Google, etc. all the models are coming from private companies. The government metastasized and seems to be unable to do anything besides create more pointless public servants.
They could spent 1 trillion on this (what's 1 trillion when you spent $42 billion to not connect a single home to broadband? and what's a trillion when you're already 36 trillion in debt), I still don't think the US goverment could beat private companies. Too much bureaucracy. Too many people with the sole goal of making the government ever bigger.
Now on a positive note (I guess), I'll grant everybody that it's even less likely that AGI would from the EU bureaucracy.
This week a major US AI development flew under the radar - a US congressional commission, the U.S.-China Economic and Security Review Commission (USCC) presented their annual report to congress.
Their number one recommendation was that Congress establish a "Manhattan Project-like program dedicated to racing to and acquiring an Artificial General Intelligence (AGI) capability".
There's a lot to say about this, which I cover in the piece.
It's an awfully big logical stretch from "a report used the phrase Manhattan Project" to "AGI may be born secret." This is coming from someone who knew the relevant history/concept beforehand.
The word "may" is doing a great deal of work in your headline.
It's a good idea. I don't think secrecy is really necessitated aside from the idea that if the US government is the one spending the money, then they should reap the benefits. Ideally, I'd love to see many governments collaborate here and share the technology with each other as a result.
What is scary about this memo, and contrary to your "lets be friends" outlook, is that we are entering a geopolitical era of precisely the opposite.
The US is withdrawing from its global cop role. Russia and China are saber rattling. NATO may be dissolved by Trump, at least the American side of it. Israel and Saudi Arabia are new strange bedfellows and allies. Japan and India are vastly increasing their military spending, and Japan will likely arm itself with nuclear weapons. Sunnis and Shiites are nearing a war. Turkey wants to throw its weight around regionally. International trade is very likely to be strongly disrupted and production nearshored.
We are entering "interesting times".
The Five Eyes may cooperate. Maybe.
> What is scary about this memo, and contrary to your "lets be friends" outlook, is that we are entering a geopolitical era of precisely the opposite.
That certainly may be true! Or perhaps, the cost of not collaborating is now too great to ignore.
> The Five Eyes may cooperate. Maybe.
On this point - check out Matthew Pines on his view that over the next year the US will force allies to convert treasuries and gold to US century bonds, with the stick being a threat of being left outside the US security and technology alliance system. The conversion being a way to deal with the big overhang of US short-dated debt.
The world is changing fast.
Where do people come up with nonsense like that?
Is there a betting market for this?
I don’t care who is researching this, we won’t have AGI by 2027 and super intelligence by 2030s.
> Based on trends in AI capabilities research since GPT-2, we are on course to expect AGI by 2027. Once AGI capability is available, if labs focus on automating AI research itself, progress in AI should accelerate. If similar progress can be achieved as the phase from GPT-2 to GPT-4, or GPT-4 to AGI, we should expect Superintelligence before the end of the decade.
"born secret", maybe. But by a US government program? Lately besides corruption and inefficiencies, the government didn't have much to show. NASA go trounced hard by SpaceX.
The Kamala "broadband" project did cost $42 billion to taxpayers and not one home has been connected with this new program. In nearly four years. Zero. Meanwhile StarLink is actually connecting people.
Be it NVidia, OpenAI, Meta, Google, etc. all the models are coming from private companies. The government metastasized and seems to be unable to do anything besides create more pointless public servants.
They could spent 1 trillion on this (what's 1 trillion when you spent $42 billion to not connect a single home to broadband? and what's a trillion when you're already 36 trillion in debt), I still don't think the US goverment could beat private companies. Too much bureaucracy. Too many people with the sole goal of making the government ever bigger.
Now on a positive note (I guess), I'll grant everybody that it's even less likely that AGI would from the EU bureaucracy.
> The Kamala "broadband" project did cost $42 billion to taxpayers
It didn't, the money has not been spent yet.
Don't let facts get in the way. GP was just trying to be inflammatory and you are here ruining everything.