Sabitlenmiş Tweet
chris camp
1.8K posts
chris camp
@christophercamp
exploring freedom enhancing technologies Prev @Blockchain @Coinbase + https://t.co/Q07jJU6csB PGP: B9F6EBE795C80F09
Santa Monica, CA Katılım Mayıs 2008
4.6K Takip Edilen704 Takipçiler
@Ritholtz Hot money / momentum traders abandoning the metals trade
and maybe some profit taking from bigger holders with real long term interests, but who see the short term trend cooling
+ risk of entering a sell-everything zone if recession hits
English
We are in the midst of a hot war in the Middle East, with Oil prices kissing $120, a cooling (pre-war) labor market, and the odds of a recession ticking up from modest levels.
And despite all of this geopolitical economic chaos, Silver and Gold are both crashing.
Besides capital flow reversals and speculative unwind, can anyone explain what is driving this action?
English
@mrjasonchoi What's your timeline to AGI?
2-3yrs?
3-5?
5-10
10-20
20+
English
As an avid user of almost every AI tool available, the current trajectory to AGI seems years, not months away.
It feels a bit like teaching a new grappler how to do Jiu Jitsu. When he gets submitted, you write down how and hand the notes to the next newbie, and so on.
We’re expecting a black belt to emerge from that process.
English
@emollick I'm not sure this lasts. Look at something like the Talaas chip, where the model is embedded directly on the chip. You get *way* more efficient inference than a generalized chip. and it's 10x faster
I expect we'll see more innovations like this
taalas.com/the-path-to-ub…
English
Yes, this is the competitive advantage argument economists have been discussing for awhile, but it is coming true.
There will not be enough compute for many years to automate many human jobs, even assuming AI can do that work and companies are willing to replace the people.
English
A big determinant of AI's job impact is driven by the lack of compute, especially for agentic work, which takes a lot of it. That makes AI expensive.
So companies will only want to burn compute on high-value tasks (eg coding), because, in other jobs, humans remain much cheaper.
English
I don’t think you can TACO tweet out of refineries being blown up tbh
OSINTtechnical@Osinttechnical
Tehran’s Shahran oil depot is completely consumed by a massive blaze after Israeli airstrikes hit the facility tonight.
English
@matthuang @tradexyz @HyperliquidX Honestly wild that the norm has been for markets to sleep all weekend
English
Global risk expressed via oil prices 24/7
cc @tradexyz @HyperliquidX
English
@ramez If current AI techniques start to hit a wall, watch avalanches of money go into efforts to dynamically map the brain
That timeline could speed up dramatically. Maps that might have taken 10-25yrs could be done in 5-8.
English
@ramez but maybe it's 1820 for AI right now
in which case we've got a good runway of growth ahead
--
In terms of closing that gap on the human brain's algorithm, I would point to the kind of work that Ed Boyden is doing.
be.mit.edu/faculty/ed-boy…
English
Here's an excellent example of those diminishing returns in R&D. Applying a lot more cognitive power - and a lot more money - did advance physics, but dramatically sub-linearly. We should expect the same in AI and be dubious of recursive self-improvement.
Roko 🐉@RokoMijic
Many people don't understand just how brutal diminishing returns in theoretical physics were. Physics barely existed before 1820. After 1970, there was essentially nothing left to discover. In 1819 there were probably less than 100 full-time paid physicists in the whole world. By 2026 there are probably about a million physicists across academia and industry, and that number was already huge in the 1970s when physics sort of "ended" with QCD and electroweak unification. A small, brave band of gentlemen-scholars and amateurs worked out the most important parts of physical law in the 1800s. People doing it as a hobby! Today, vast armies of professionals equipped with supercomputers toil away in the quantum gravity dungeon, unable to make progress. Diminishing returns are brutal.
English
@emollick "We're gonna need a bigger chart.."
If we've already hit 3x this fast, it's not crazy to think we could hit 25x or 100x in 10yrs. Even 1,000x may be in reach.
That is nuts. and potentially massively disruptive in terms of jobs/econ even if we don't hit AGI
English
Given the GDPval benchmark for GPT-5.4, I've updated this chart, the new model ties or beats humans as judged by other experts at professional tasks 82% of the time
If you give a 7 hour task to AI, even with failure rates and the need to check results, you'd save 4h 38m average
Ethan Mollick@emollick
Since OpenAI didn't update Figure 7 from GDPval given the success rate of GPT-5.2 on long-form tasks, I used GPT-5.2 Pro to do so. The chart assumes the process is: delegate long tasks to AI, evaluate the output for an hour, then decide to try again or give up & do it yourself.
English
@Econ_4_Everyone - You might be right in your assessment, but there's also a non-zero chance that you are wrong. And we should still spend energy preparing for that non-zero scenario, for a world with very, very capable AIs (but still short of AGI).
English
@Econ_4_Everyone (and their predictions have been fairly accurate so far)
- Very recent models (like claude code) signal a step change relative to earlier models
- The entire world is racing to solve intelligence - trillions being spent and planned. Hundreds of startups attacking every angle
English
When AI first arrived on the scene, I worried it would make economists, or even critical thinkers more broadly, less valuable. In my travels in the past 6 months to work with non-profits, for profits, and government agencies, I have observed how people are actually using AI. I have watched them fumble around with insights they clearly did not create themselves.
My fears are now assuaged. One observation is that AI can produce something that in some cases is very wrong and in others looks nearly right, but is not quite there. Even if in time AI improves to "nearly right" or "exactly right" every time, a second issue still arises: explaining the materials.
Explaining why an answer is almost correct but subtly off requires exactly the critical thinking skills that created the knowledge in the first place. Even explaining "exactly right" material takes critical thinking. I've watched smart people confidently present AI-generated material they clearly don't fully understand. The words sound right. But when someone pushes back just a little bit, the sand castle crumbles.
It is quite difficult to defend what you didn't build. This leads me to now make the optimistic case for human expertise. The value of deeply understanding something — of having built the knowledge yourself — hasn't diminished with AI. If anything, it's increased. The people who can tell the difference between "nearly right" and "right" are more valuable than ever. The people who can explain the subtle details about something that is exactly right are invaluable.
Creating knowledge still matters. Maybe now more than ever.
English
@urbit time for this
x.com/aidaxbaradari/…
Aida Baradari@aidaxbaradari
Today, we're introducing Spectre I, the first smart device to stop unwanted audio recordings. We live in a world of always-on listening devices. Smart devices and AI dominate our world in business and private conversations. With Deveillance, you will @be_inaudible.
English
There is a growing conflict, brewing beneath the surface of our digital lives.
Historically, your 'recording devices' were your biological eyes and ears. Your storage was your own memory.
Sure, if you were a dissident or a foreign diplomat you may have been a bit more savvy to data collection practices. But even then, such experiences have only arisen in the last hundred years, if that.
On an evolutionary timescale, this is nothing.
But now in the age of cheap digital devices and increasingly abundant AI inference, comprehensive data collection is becoming compelling for the average person. And for wholly innocuous use cases.
"I want to record our weekly 1:1 so I can collate useful action items after the fact while staying focused on the conversation"
"I want to keep track of the things I said I would do throughout the day so I can follow through on the promises I make to my friends"
This desire drives a demand. And also a backlash...
x.com/ycombinator/st…
English
chris camp retweetledi
you see a toddler pass out mid-morning, gripping a tiny half eaten muffin, and when they wake up later, they just start eating the muffin again. this is real freedom. there are wild horses less free than this
English
@Evan_ss6 uncanny insights
unparalleled returns
can't find this quality anywhere else. and that's why 9 out of 10 doctors recommend taking evanss6 each night before bed
English
@TorBair @kevinroose This might be the most important question in all of this
We have so little transparency
Pull the curtain back and Dario and Hesgeth could be having a laugh over their manufactured kerfuffle
It's useful to remember: We just don't know
English
@kevinroose To what extent do you think the public statements from each side reflect the content of the *actual* private conversations vs posturing designed for maintaining leverage and eventual face-saving?
English
Agree with him or not, the (oddly popular on here!) take that Dario Amodei is some kind of bumbling Silicon Valley naïf who couldn't get a deal with the Pentagon done because he doesn't understand politics seems entirely wrong.
His favorite book is "The Making of the Atomic Bomb." He used to buy copies for new Anthropic employees. (There's still a copy prominently displayed in the Anthropic library.) He fully expected -- back when it was a crazy thing to expect! -- that AI would become as important as nuclear weapons, and that the people who built it, like the scientists of the Manhattan Project, would face pressure from governments to use their technology in ways they found immoral or dangerous.
I am sure this all could have been handled differently, but none of this is a surprise to anyone who knows anything about the relevant people.
English
chris camp retweetledi
Think about the power Hegseth is asserting here. He is claiming that the DoD can force all contractors to stop doing business of any kind with arbitrary other companies.
In other words, every operating system vendor, every manufacturer of hardware, every hyperscaler, every type of firm the DoD contracts with—all their services and products can be denied to any economic actor at will by the Secretary of War.
This is obviously a psychotic power grab. It is almost surely illegal, but the message it sends is that the United States Government is a completely unreliable partner for any kind of business. The damage done to our business environment is profound. No amount of deregulatory vibes sent by this administration matters compared to this arson.
Secretary of War Pete Hegseth@SecWar
This week, Anthropic delivered a master class in arrogance and betrayal as well as a textbook case of how not to do business with the United States Government or the Pentagon. Our position has never wavered and will never waver: the Department of War must have full, unrestricted access to Anthropic’s models for every LAWFUL purpose in defense of the Republic. Instead, @AnthropicAI and its CEO @DarioAmodei, have chosen duplicity. Cloaked in the sanctimonious rhetoric of “effective altruism,” they have attempted to strong-arm the United States military into submission - a cowardly act of corporate virtue-signaling that places Silicon Valley ideology above American lives. The Terms of Service of Anthropic’s defective altruism will never outweigh the safety, the readiness, or the lives of American troops on the battlefield. Their true objective is unmistakable: to seize veto power over the operational decisions of the United States military. That is unacceptable. As President Trump stated on Truth Social, the Commander-in-Chief and the American people alone will determine the destiny of our armed forces, not unelected tech executives. Anthropic’s stance is fundamentally incompatible with American principles. Their relationship with the United States Armed Forces and the Federal Government has therefore been permanently altered. In conjunction with the President's directive for the Federal Government to cease all use of Anthropic's technology, I am directing the Department of War to designate Anthropic a Supply-Chain Risk to National Security. Effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic. Anthropic will continue to provide the Department of War its services for a period of no more than six months to allow for a seamless transition to a better and more patriotic service. America’s warfighters will never be held hostage by the ideological whims of Big Tech. This decision is final.
English
@SGJohnsson Motivated cognition doing a lot of heavy lifting in @PalmerLuckey's comparison
As a matter of principle, DoW's approach is v concerning. Add in the reality that the tools it wants total control over are getting more powerful by the day and alarm bells should be going off
English
chris camp retweetledi
We’re analogizing a weeks-long taking of physical infrastructure to avert a LABOR dispute 80 years ago to a permanent deputizing of a private company’s people and resources for purposes of killing and surveillance.
What are we doing here?
If you want to know why DoW prefers to be solely limited to “legal” applications is because, as John Yoo made abundantly clear following 9/11, you can always find a lackey in OLC to draft a memo in secret under some demented legal theory to give cover for anything, including torture and dragnet surveillance of Americans (oh, is that what we’re talking about here…?)
Palmer Luckey@PalmerLuckey
"The President can't make private companies work with the military, it would be unprecedented!" May 10, 1948 I HAVE today by Executive order taken over the country's railroads and directed the Secretary of the Army to operate them in the name of the United States Government.
English