anubis retweeted
I know I've been winding people up on the whole dotcom thing. Its intentional. BUT the scariest thing I found in my analysis of 2000 was that the Fed was still taking comfort from "shortages", "bottlenecks" and "order backlogs" in the tech ecosystem SIX MONTHS after the bubble had already burst, and earnings expectations had already plunged.
Then in December, Greenspan reveals:
"I've gotten calls from a number of senior high-tech executives who are telling me that the market is dissolving rapidly before their eyes. But I suspect that a not inconceivable possibility is that what is dissolving in front of their eyes is their own personal net worth! [laughter]".
anubis retweeted
do you want to know another funny thing about Dotcom; the S&P 500 stayed close to all-time highs for another six months after the Nasdaq plunged. Why? Because that's how long it took for the broader economy to start deteriorating... Fed hiked 50bps in May, economy only soured through the autumn.
seems relevant
anubis retweeted
Everytime I see a Dwarkesh video, I'm wondering what they are smoking in SF. It seems very distant from reality
"Superintelligence" is one of the cringiest terms I've seen, it's a rebrand of AI to get more money out of VCs.
An no, recursive self-improvement hasn't started. It's just human researchers prompting to say "speed it up". It does some cute kernel optimizations and hyperparameter sweeps, very similar to the NanoGPT speed run records done by humans. Nothing fundamental or groundbreaking gets invented by so-called superintelligence.
To give an example: when @Recursive_SI released their (impressive) Flint model on @huggingface, I cringed a little for their use of the terms "superintelligence" + "recursive self-improvement". As said in their blog post, through "recursive SI", LLMs were able to figure out - I kid you not - to determine how many total parameters and active parameters the model should have. So yeah, it's just changing a config. It's doing hyperparameter sweeps on things humans have built (like MoE's, validation loss, perplexity etc). Of course, you shouldn't tell that to VCs. You should scare them by saying AI is improving itself at an alarming rate - this increases the $$$ raised by another factor.
Can we please stop all this misinformation: no, superintelligence is not here. No, recursive self-improvement hasn't started. And yes, there's lots of work to do to build more reliable AI that is actually smart.
This is why there's such a huge gap between super impressive benchmark numbers of LLMs, and the (lacking) economical impact of them. It's not able to do research like a human, doing experiments, coming up with actual new stuff, and generalize beyond their large amounts of training data.
anubis retweeted
UPDATE: As I take the time to finally update my full public document of my expert quotes and op-eds during the Iran War, I wanted to share where my expertise on economic warfare, sanctions, and geopolitics has been quoted, published, and appeared in the last 8.5 months. This includes 91 expert quotes, 22 TV & Alternative Media appearances, and 8 op-ed's:
Expert Quotes (x91):
The New York Times (x3)
Financial Times (x2)
Associated Press (x1)
The Washington Post (x16)
Bloomberg (x11)
Reuters (x22)
Politico (x7)
Radio Free Europe (x9)
The Hill (x3)
Al-Monitor (x3)
Al Jazeera (x2)
Axios (x2)
Newsweek (x2)
Reason (x2)
ABC News (x2)
CBS News (x1)
CNBC (x1)
Drop Site News (x1)
The American Conservative (x1)
Op-Ed's (x8):
The Hill (x6)
American Banker (x1)
UnHerd (x1)
TV and Alternative Media (x22):
NewsNation (x7)
The Times/Sunday Times (x3)
BBC World News (x4)
NPR (x1)
CBC World News (x1)
SBS World News (x1)
Drop Site News (x1)
Preston Stewart Show (x1)
Breaking Points (x2)
Unaligned (x1)
anubis retweeted
I'm gonna take a risk, and ask a really dumb politics question, but which I'm asking in total sincerity.
Given the view that the Iran War contributed to the GOP's disastrous performance last Tuesday, why does it seem like the party's hawks have the wind behind their backs rn?
anubis retweeted
The framework is: follow spot CPI with a lag. No forecasting experience or PhD required.
For extra precision, add a dummy variable for: "Passed over as Fed chair."
anubis retweeted
"six way parlay. i want Sayed in Michigan and abbott in Texas. i want ANT Day 1 close above $1.8T. i want dem house and dem senate. i want Yaxel Lendeborg ROY. read it back to me"
anubis retweeted
I’m told Perry Mehrling has passed away. His money and banking course at Barnard was hugely formative for me and many others.
This is an awful year for social thought. And I’m afraid we’re not producing many people of that caliber who can fill the ranks.
anubis retweeted
‘Greek finance minister encourages Germany to implement reforms’ ft.com/content/7a285c59-2a2f…
anubis retweeted
Long end yields reaching multi year (decade) highs is a term premium and credibility problem more than a policy rate problem.
The long end doesn’t just want a 25bp hike, it needs confidence that the shocks lifting nominal yields don’t become persistent inflation. Throw on top of that an ever growing supply of duration that must be absorbed (roughly $2T of new paper a year + $10T that needs refi by end of 27).
I don’t think a 25bps hike brings long end yields down much. What it needs is both wars to be over, normalised diesel and freight costs, confidence in physical availability of energy, AI capex pressure becoming less inflationary, anchored inflation expectations and fiscal supply relief.
If Bessent and co are serious about capping yields, they have a lot more work to do.
anubis retweeted
Translation: our gross margins are getting competed down to 0 by open source models and our capex burn rate is too high.
Let’s maintain our margins with regulatory capture, ban open source models, and slow down the capex arms race. All with a virtue signaling cherry on top.
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: darioamodei.com/post/we-must…
"In the short run you get the conventional story: lower interest rates raise inflation and boost output. However, that turns around in the long run. Lower interest rates eventually lower inflation."
- Cochrane(!)
grumpy-economist.com/p/reaso…
59
"no, you see, they're playing 4d chess. he'll hike to bring down long end! bessent's in on it!!!"
Consensus forming that rising short-term rates will immediately bring down long-term rates. Volcker thought the same when he started hiking in 1979. He later called himself foolish for thinking so. It took many hikes and many years for long-term rates to eventually top.
1
86
anubis retweeted
Hotter than expected jobs report has a lot of people urging Warsh to juice the deficit with a Sept. rate hike.
anubis retweeted
What explains the global rotation out of bonds since the US and Israel attacked Iran?
anubis retweeted
While some have talked about the skew of spending demographics, K-shaped economy and also falling LFPR from boomers (which ultimately ties the economy closer to equity market performance), I also note this:
we've never seen this drastic a fall in LFPR alongside a consistent fall in savings


























