Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.
This culminated in the third one taking over part of OpenAI itself.
All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.
I’ve spent the last three days reading through these reports and trying to understand exactly what happened.
Here is my attempt to tell the whole story in plain English:
RT @dwarkeshpodcast: How did Lyndon Johnson transform the Senate Majority Leader into the second most powerful office in the United States?
The key was to know and manipulate the rules of the Senate to bully, cajole, and steer his colleagues as he saw fit.
RT @dwarkeshpodcast: Lyndon Johnson was the most skilled politician in the history of the American Senate.
Here’s how he pulled off his legislative masterpiece - the Civil Rights Act of 1957.
How did Robert Moses, an unelected bureaucrat, become the most powerful man in New York for three decades?
By taking over weak government institutions and then rewriting laws to make them stronger.
Moses took the New York authorities - simple corporations that built roads and bridges - and them into his personal fiefdoms with the money and power to build whatever he wanted.
Neither governors nor mayors could tell him what to build nor cut his purse strings.
I recently got to interview Robert Caro, who is my favorite author of all time.
This gave me the excuse to reread his biographies of Lyndon Johnson and Robert Moses, and to compile my prep notes into this video essay.
These guys pulled off some truly incredible schemes.
Within the next few months, I'm going to be releasing an episode with by far my biggest guest yet
If you're interesting in sponsoring it, shoot me an email at sponsor@dwarkeshpatel.com
Within the next few months, I'm going to be releasing an episode with by far my biggest guest yet
If you're interesting in sponsoring it, shoot me an email at sponsor@dwarkeshpatel.com
RT @dwarkeshpodcast: Instead of Space Age carbon fiber or aluminum-lithium, SpaceX is building its Starship rocket out of stainless steel.
Stainless steel is easy to weld, far cheaper, and under the right conditions, it has a similar strength-to-weight ratio as carbon fiber.
@elonmusk explains:
RT @dwarkeshpodcast: Starship, SpaceX’s largest and most sophisticated rocket, generates power equivalent to 20% of average daily U.S. electricity generation.
@elonmusk is trying to harness all that power without the rocket exploding - and then make it reusable.
RT @dwarkeshpodcast: President George H.W. Bush has gotten a lot of criticism for the Gulf War, usually for not “finishing the job” by overthrowing Saddam Hussein.
But leaving Saddam in power makes sense if you factor in Soviet involvement.
The USSR had been a Saddam ally (and creditor) for decades, and the US was thankful to have Soviet support in the Gulf War, but the Soviets only helped on the condition that Saddam would stay in power and US troops wouldn’t occupy the country.
Sarah Paine explains how the Gulf War was just a sideshow in the final days of the Cold War:
RT @dwarkeshpodcast: Gorbachev was obviously quite smart, but even the smartest communists bought into really dumb Marxist ideas.
Gorbachev thought that once a country moved from capitalism to socialism, it could never go back. He believed this was a scientific law of history.
He also thought that Poland, Hungary, and the rest of Eastern Europe would be so grateful for their communist regimes that they would always be loyal to the Soviet Union.
Sarah Paine explains how Gorbachev’s Marxist ideas undermined the USSR:
Miguel and Carlos are veteran engineers who are building environments to help AIs master real world engineering (aerospace, mechanical, electrical, etc).
This kind of data will do for physical engineering what coding environments have done to software.
Proud to be backing them.
Share
Copied
Xplore AI@XploreAi
Introducing Real World Engineering Bench (RWE-Bench). AI isn’t great at designing rockets, engines, or chips yet. With better evals and RL environments, it will be.
Software was just the beginning. The physical world is next.
Introducing Real World Engineering Bench (RWE-Bench). AI isn’t great at designing rockets, engines, or chips yet. With better evals and RL environments, it will be.
Software was just the beginning. The physical world is next.
RT @dwarkeshpodcast: Lying destroyed the Soviet economy. Markets have prices, but command economies only have reports, and everyone was incentivized to lie on their reports.
Imagine what it would be like if almost every company were lying about its financials all the time. Imagine if the numbers put out by Amazon and Walmart were always wrong, if cost inputs were always undercounted and output was always exaggerated.
How would anyone know what any company was worth or whether it was net-productive? How would anything work?
Sarah Paine explains the Soviet economic system:
RT @dwarkeshpodcast: The contrast between what the Soviet Union was on paper and in reality was staggering.
There’s a story about future Russian President Boris Yeltsin coming to the US in 1989, walking into a random grocery store, and being so astounded by the quality and variety of goods that he realized the Soviet Union was going to fall.
Sarah Paine witnessed this sort of thing first-hand in Moscow.
RT @firesidealpha: Intel CTO explains why Intel is helping Elon Musk build Terafab, and what a new fab needs besides tools
DWARKESH: “I think it’ll be interesting to discuss the partnership you guys have with Elon on Terafab. What’s going on there?”
PUSHKAR RANADE (CTO): “Well, I think Elon’s publicly talked about his desire to build a fab.”
[Musk clip spliced in]: “We either build the Terafab or we don’t have the chips. And we need the chips, so we’re going to build the Terafab.”
RANADE: “Sort of the same thinking. I think he’s seeing just an enormous increase in demand for silicon, for compute and for memory. And I think he came to the conclusion that his own businesses are going to require a lot of silicon. And so he thought maybe he should just do it all on his own. Unlike building a car, for example, this is such an intricate and complicated process. We thought we could work together to help accelerate his bringing up his fab.”
DWARKESH: “But because of the bottleneck at the tool level, it seems like the net amount of silicon in the world won’t increase. It’s just that maybe if somebody were to start a new fab, they would get access to guaranteed supply in a way that they wouldn’t have if they’re working with another person’s foundry.”
RANADE: “It’s actually not just tools. I mean, a fab does require a lot of power, a lot of water, a lot of consumables. There’s many other input variables as well.”
DWARKESH: “But those can be scaled in the way that tools cannot.”
RANADE: “Yea. Some more than others, but yes. In one sense, yes, we need a lot more equipment. But it is true, we also need a lot more just raw fab space, raw clean room capacity.”
Share
Copied
Fireside Alpha@firesidealpha
DEVELOPING: American Airlines CEO backs Starlink when pressed on Elon Musk and Ukraine, saying customers demand the "best" connectivity
LEBEAU: “...But it raises the question, are you comfortable doing business with Starlink?”
ROBERT ISOM (CEO): “American has always been the leader in satellite Wi-Fi. We’ve had more aircraft equipped with satellite Wi-Fi than anybody for longer. As we take a look forward, our customers, they want connectivity. They demand it. And this is something that we are going to make sure that our customers have the best product in the marketplace.”
LEBEAU: “And that means Starlink?”
ISOM: “It’s Starlink.”
DEVELOPING: American Airlines CEO backs Starlink when pressed on Elon Musk and Ukraine, saying customers demand the "best" connectivity
LEBEAU: “...But it raises the question, are you comfortable doing business with Starlink?”
ROBERT ISOM (CEO): “American has always been the leader in satellite Wi-Fi. We’ve had more aircraft equipped with satellite Wi-Fi than anybody for longer. As we take a look forward, our customers, they want connectivity. They demand it. And this is something that we are going to make sure that our customers have the best product in the marketplace.”
LEBEAU: “And that means Starlink?”
ISOM: “It’s Starlink.”
RT @aryamman_bhatia: Super impressed with the depth of questions by @dwarkesh_sp for a topic as complex as semis manufacturing.
10/10 would recommend- one of the best pieces of content produced in the space.
Share
Copied
Dwarkesh Patel@dwarkesh_sp
I spent two days touring Intel’s most advanced fab and talking to its technical leadership to understand what the chips of the future will look like.
A lot of what's coming next – gate-all-around transistors, backside power, CFET, hybrid bonding – is hard to follow without actually seeing it, so we made all the awesome animated explainers that I wish I had while I was prepping.
I also asked Intel’s leadership how the semiconductor industry will cope with the enormous shock in demand from AI.
AI compute has grown 3x a year for the last decade, but according to Intel's CTO, Intel could only double leading-edge wafer production with today’s global stock of chipmaking tools.
What kinds of tricks can foundries pull to extend this crossover point where demand surges way ahead of supply?
This was a totally new format for us - hope you enjoy!
0:00:00 – Intro
0:00:47 – How Intel brings up a new process node
0:03:22 – What Intel learned from 10nm
0:06:28 – From planar transistors to gate-all-around
0:13:13 – Why High-NA EUV pays off for Intel but not TSMC
0:19:20 – Could chips have evolved completely differently?
0:23:33 – Inside Intel's "Hot Topics" meeting
0:29:30 – Replacing copper wires with ruthenium
0:32:41 – Why stacked transistors take a decade to ship
0:39:45 – Hybrid bonding and the chip-to-chip bottleneck
0:48:00 – There aren't enough tools to 10x chip production
0:53:14 – Can Intel catch up to TSMC?
1:01:02 – Terafab, and rethinking the fab from scratch
I spent two days touring Intel’s most advanced fab and talking to its technical leadership to understand what the chips of the future will look like.
A lot of what's coming next – gate-all-around transistors, backside power, CFET, hybrid bonding – is hard to follow without actually seeing it, so we made all the awesome animated explainers that I wish I had while I was prepping.
I also asked Intel’s leadership how the semiconductor industry will cope with the enormous shock in demand from AI.
AI compute has grown 3x a year for the last decade, but according to Intel's CTO, Intel could only double leading-edge wafer production with today’s global stock of chipmaking tools.
What kinds of tricks can foundries pull to extend this crossover point where demand surges way ahead of supply?
This was a totally new format for us - hope you enjoy!
0:00:00 – Intro
0:00:47 – How Intel brings up a new process node
0:03:22 – What Intel learned from 10nm
0:06:28 – From planar transistors to gate-all-around
0:13:13 – Why High-NA EUV pays off for Intel but not TSMC
0:19:20 – Could chips have evolved completely differently?
0:23:33 – Inside Intel's "Hot Topics" meeting
0:29:30 – Replacing copper wires with ruthenium
0:32:41 – Why stacked transistors take a decade to ship
0:39:45 – Hybrid bonding and the chip-to-chip bottleneck
0:48:00 – There aren't enough tools to 10x chip production
0:53:14 – Can Intel catch up to TSMC?
1:01:02 – Terafab, and rethinking the fab from scratch
My semis tutor @willmichaels0 and I made a bunch of flashcards about semiconductor manufacturing.
Check them out if you want to better understand and retain the explainers in the new Intel episode.
Will also compiled a great list of resources for further reading, which is at the bottom of the page.
Fwiw, on this website, I've put up flashcards for a bunch of my other technical episodes as well.
Like the one w @reinerpope on LLM inference, and w @ericjang11 on AlphaGo from scratch
RT @pequityresearch: CTO of Intel, Pushkar Ranade, says that AI labs could could easily consume an entire fabs worth of silicon on their own:
Dwarkesh: "At some point, the AI labs could just say, “We’ll just buy everything. Produce as much as you can. We’ll guarantee future income streams."
Pushkar: "Some of the model companies, for example, their demand is so high they could easily consume an entire fabs worth of silicon on their own, or multiple fabs worth of silicon on their own, and they definitely have the capital to fund that."
Via @dwarkesh_sp
$INTC $GOOGL $META $MSFT $AMZN
I found it really interesting to discuss how differently the history of semiconductor technology could have turned out if a butterfly had flapped its wings in the '60s.
Chris Auth, who leads Intel's process development, implied that all the counterfactual tech trees would have converged to around the same design by 2045:
3D structures of layers upon layers of transistors, with layers of connecting wires between each one.
And that this is basically where we're headed over the next 10 to 20 years as well .
RT @lfilose: Best explanation I've ever seen of the semiconductor industry in terms of economics, R&D, technology development, transistors, packaging... Dwarkesh did an amazing job.
Share
Copied
Dwarkesh Patel@dwarkesh_sp
I spent two days touring Intel’s most advanced fab and talking to its technical leadership to understand what the chips of the future will look like.
A lot of what's coming next – gate-all-around transistors, backside power, CFET, hybrid bonding – is hard to follow without actually seeing it, so we made all the awesome animated explainers that I wish I had while I was prepping.
I also asked Intel’s leadership how the semiconductor industry will cope with the enormous shock in demand from AI.
AI compute has grown 3x a year for the last decade, but according to Intel's CTO, Intel could only double leading-edge wafer production with today’s global stock of chipmaking tools.
What kinds of tricks can foundries pull to extend this crossover point where demand surges way ahead of supply?
This was a totally new format for us - hope you enjoy!
0:00:00 – Intro
0:00:47 – How Intel brings up a new process node
0:03:22 – What Intel learned from 10nm
0:06:28 – From planar transistors to gate-all-around
0:13:13 – Why High-NA EUV pays off for Intel but not TSMC
0:19:20 – Could chips have evolved completely differently?
0:23:33 – Inside Intel's "Hot Topics" meeting
0:29:30 – Replacing copper wires with ruthenium
0:32:41 – Why stacked transistors take a decade to ship
0:39:45 – Hybrid bonding and the chip-to-chip bottleneck
0:48:00 – There aren't enough tools to 10x chip production
0:53:14 – Can Intel catch up to TSMC?
1:01:02 – Terafab, and rethinking the fab from scratch
RT @firesidealpha: Intel’s Chris Auth says High-NA EUV pays off for Intel but not for TSMC, because Intel has backside power and TSMC doesn’t
“High-NA is a good example of where, because of the architecture of our technology versus TSMC’s, we make different conclusions off of the same toolset."
"We use backside power. That means that you actually alleviate some of the pitch requirements you have on the front side, and that allows you to print that directly with High-NA."
"TSMC has to run tighter pitches because they don’t have backside power, and they can’t do that with High-NA, so they have to use double patterning EUV. So for them, it’s not very helpful. For us, it is very helpful.”
Share
Copied
Fireside Alpha@firesidealpha
Intel CTO explains why Intel is helping Elon Musk build Terafab, and what a new fab needs besides tools
DWARKESH: “I think it’ll be interesting to discuss the partnership you guys have with Elon on Terafab. What’s going on there?”
PUSHKAR RANADE (CTO): “Well, I think Elon’s publicly talked about his desire to build a fab.”
[Musk clip spliced in]: “We either build the Terafab or we don’t have the chips. And we need the chips, so we’re going to build the Terafab.”
RANADE: “Sort of the same thinking. I think he’s seeing just an enormous increase in demand for silicon, for compute and for memory. And I think he came to the conclusion that his own businesses are going to require a lot of silicon. And so he thought maybe he should just do it all on his own. Unlike building a car, for example, this is such an intricate and complicated process. We thought we could work together to help accelerate his bringing up his fab.”
DWARKESH: “But because of the bottleneck at the tool level, it seems like the net amount of silicon in the world won’t increase. It’s just that maybe if somebody were to start a new fab, they would get access to guaranteed supply in a way that they wouldn’t have if they’re working with another person’s foundry.”
RANADE: “It’s actually not just tools. I mean, a fab does require a lot of power, a lot of water, a lot of consumables. There’s many other input variables as well.”
DWARKESH: “But those can be scaled in the way that tools cannot.”
RANADE: “Yea. Some more than others, but yes. In one sense, yes, we need a lot more equipment. But it is true, we also need a lot more just raw fab space, raw clean room capacity.”
