Every Day I’m Galerkin (@BikeMath) header
Every Day I’m Galerkin (@BikeMath) profile picture

Applied Mathematician, Veteran, Nevadan — Probably watching Star Trek. (he/him/his)

Las Vegas, NV
Joined April 2010
Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)Photo by Every Day I’m Galerkin (@BikeMath)
Every Day I’m Galerkin retweeted
"Who should Senate Democrats choose to control the senate?" Someone Else: 57% Schumer: 10% YouGov Blue / Nov 9, 2026
2
147
1
1,364
13,532
Every Day I’m Galerkin retweeted
The least credible polite fiction in America is that Schumer wants to end the Iran war. Literally no one thinks this. It is second only to the Lakeshore Drive 40mph speed limit polite fiction that no one obeys & if you try to obey it they run you off the road and kill your family
5
82
3
711
11,250
Every Day I’m Galerkin retweeted
Jensen Huang packing his things into a box, going to the lobby, and trying to call an Uber home, then remembering with dawning horror that he doesn’t know his home address
15
1
442
6,551
Every Day I’m Galerkin retweeted
The future is so bleak oh my god
Microsoft has filed a patent for an AI system that could rewrite what people say during livestreams in real time. It could change words depending on who's watching, taking into account their age, location, and language. The idea is to clean up broadcasts by removing filler words, censoring swearing, or even replacing certain words. The AI could also adjust the speaker's lip movements so viewers wouldn't notice anything had been changed.
39
1,486
9
13,280
102,282
Every Day I’m Galerkin retweeted
The “we let Israelis sexually assault kids and go back to Israel instead of facing trial” caucus has spoken I guess Vote Aaron ford
Clark County District Attorney Steve Wolfson broke party lines to endorse Gov. Joe Lombardo’s re-election campaign because of his relationship with Lombardo and his gripes with Attorney General Aaron Ford, he told the Review-Journal. FULL STORY: lvrj.com/post/3907347
7
24
1
260
12,775
Every Day I’m Galerkin retweeted
If you think this is about “hurt feelings” sry you’re a child. Stevens has been making veiled threats (of, at minimum, un-endorsing) for weeks unless El Sayed denounced ALL his meaningful criticisms of israel. He refused. At some point AIPAC had to follow through on these threats
NEW The New York Times, which spoke to Haley Stevens, basically reports that this was all about hurt feelings. So many people today made this all about Stevens putting country over party. “On Sunday morning, Ms. Stevens woke up angry about being told not to come to church. She called Mr. Rogers and asked if they could attend church together, a gesture he immediately embraced, she later explained. She told none of her friends and allies in Democratic politics about her plans.” Full Story: nyti.ms/47UNnid
12
208
5
2,007
39,286
Every Day I’m Galerkin retweeted
No disrespect to Prof. @AlexKontorovich. This is however exactly what I have been warning you about. It takes guts to admit that we are training people today for jobs that used to exist 55 years ago. And that is what got said. Remember what I am claiming: top students used to survive in mathematics. See the graphics attached from top advisors in the earlier era when jobs existed...the students survived! It was never as Alex portrayed it to be. No current advisor anywhere has these kind of survival statistics. That is what is being covered up. And there are MANY MANY more advisors who shouldn't be advising a single PhD student. Ever. We were training our future colleagues. Now we are exploiting workers misclassified as students so we can import them in massive numbers and work them in a plantation system. The students generally don't survive. And then we lie about them and say that they weren't good enough! That is literally how the system was rearchitected when the jobs stagnated over 55 years ago. And the sharp elbows edge out the sharpest minds often in the academic hunger games. Literally that is what happened. That's why the greats mostly dried up. Now we have *worse* mathematicians on average who are more incrementalist survivalists than romantic heroes and we are stuck pretending that there has not been a massive decline from the Grothendieck, Mllnor, von Neumman, Atiyah, Serre era. I am not blaming Prof. @AlexKontorovich. He has survived and I salute that. But make no mistake, he never saw the system that actually worked. I barely did when I was around 20 and it was already in flames. But unlike Alex, I was often in the room in the 1990s in Washington DC when the post-cold war sausage was being made at NSF and on the Hill. And what we did was to largely destroy the university system. Read what he writes. He is talking about the Hunger Games where one of the most important mathematicians, Zhang, worked at @subway and that is just fine. It's like training highly specialized neurosurgeons to be baristas or call center operators. It just doesn't make any sense and the mathematics community does not notice that it is operating a planation where it has to dispose of bodies and train our geopolitical rivals in sensitive subjects to stay afloat. I am sorry. But we went mad under Reagan and then insane under Clinton. This actually makes no sense now and made no sense then when I tried to stop it. It's a suicidal plantation economics operation that makes us weak and our geopolitical rivals strong. Full Stop. That's all there is to it.
In the US, we graduate on the order of 1000 math PhDs a year, and there are about 300 postdoc jobs (and ~70 tenure track jobs three years later). So no, a math PhD is not, and has never been, a "gold medal" (in the sense of an academic position). And is "merit" about potential or about actual results? I really hope it's not the former (which is much more error prone to measure)! About Zhang specifically, there's much more to his story than what's been reported, and it's devastating how much he suffered before his breakthrough.
57
176
19
1,472
152,485
Every Day I’m Galerkin retweeted
Terance Tao: Imagine prescribing a drug no one understood. Medical community:
112
889
82
12,970
293,390
Every Day I’m Galerkin retweeted
I disagree with Terence Tao's view of mathematics, which he articulated in a talk yesterday at Caltech (see slide below). Applied math problems are not disjoint from pure math problems. This is not a minor matter. So no, this is not a "more accurate model". It's embarrassing.
34
36
15
375
48,300
Every Day I’m Galerkin retweeted
Update: Hundreds of papers published (the package) ✅ The author list is "the consortium" ✅ Raw data not released ✅ Within days errors found in multiple publications ✅ One person claims credit for all the work of the consortium and compares himself to Newton ✅
Mathematics has now fully morphed into biology: Project costs millions $ ✅ Put out paper that nobody has read ✅ Massive paper supplement nobody will read ✅ Advertise with pretty art ✅ Vicious authorship fight ✅ Say you're curing cancer ✅
12
131
7
1,467
82,886
Incredible y axis. No notes. Perfect bit
Open Al's internal benchmarks show GPT-6.1 Sol crushing Opus 5.5, with Anthropic struggling to keep up
20
850
5
26,572
511,281
Every Day I’m Galerkin retweeted
uhh, let me be clear: better things aren't possible
Former President Obama says Democrats made a mistake by making goals materialistic and instrumental: "What people are starved for at the moment is the ability to tell a moral and ethical story about how we live together in this moment."
21
232
1
3,701
47,681
Every Day I’m Galerkin retweeted
I forbid my children to be smart. That is a 20th century mindset, Luddite behavior.
6
27
450
10,228
Every Day I’m Galerkin retweeted
Work is proceeding on the Hodge conjecture.
1
3
3
94
12,063
Every Day I’m Galerkin retweeted
Everyone in this video is sitting
Standing room only for Microsoft’s Amy Hood
1
10
691
Every Day I’m Galerkin retweeted
You're gonna die. Here's why that matters.
45
432
47
4,157
207,414
Every Day I’m Galerkin retweeted
I'll bet there are other mathematicians scooped by OAI in the past few months that are going to be looking back very closely at what they did or didn't put into codex
1
4
173
2,299
Every Day I’m Galerkin retweeted
Paul Atreides: Dune 1 vs Dune 3
17
1,154
18
21,122
409,644
created theranos as a trauma response
honestly kind of gross to paint a woman who was raped and groomed as some sort of liar or sociopath when she clearly created theranos as a trauma response (and motivated by Balwani). is she innocent? no, ofc not. but it feels a little weird to paint her as some villain.
5
196
6,873
107,464
Every Day I’m Galerkin retweeted
SemiAnalysis' Jordan Nanos reveals the OpenAI engineers scrolling their own kernel code had no idea what it did line by line, and it didn't matter because the AI does "It's not slop, because it performs, but it's literally just AI-generated assembly basically puked out in Gluon, this low-level kernel programming language that they built on top of Triton, which uses this really interesting programming model." "The point is, the guys who were scrolling through this code with us, it was kind of clear that they know a whole bunch about hardware, they know a whole bunch about the concepts in the system, and they just have no idea what this MLA kernel that they're showing us for DeepSeek actually does." "You can go line by line and it's like nope, nope, nope." "But it doesn't matter. The AI understands it, the AI tests it, and you see the results. It produces correct kernels that perform really well." "And I think this is just a sign of what's to come again." "That actual code is not necessarily something that a human has to reason about deeply, if the AI knows how to manipulate the data movement and the processing elements on the hardware that you've given it." ________ More takeaways from OpenAI's Jalapeno development: firesidealpha.substack.com/8…
SemiAnalysis' Jordan Nanos says OpenAI beats Nvidia on both fast and cheap tokens at once, where every other challenger could only pick one side of the curve "I think the conclusion is that they're basically winning on both sides of the curve." "So both fast tokens and cheap tokens, which is just so interesting, because we've seen so many other companies make claims about how they're going to beat Nvidia and they just pick one of those." "Groq or Cerebras, or any of the other startups that are going to focus on SRAM, call it a d-Matrix that's coming up with stuff, or SambaNova, where we've even seen some results on the InferenceMAX benchmark or something similar to it." "They're saying they're going for fast tokens, just decode speed, low batch size, they don't worry about throughput." "And then you've got other guys that are worried about throughput, call it AMD just as a simple example, but even TPU or Trainium could be in this bucket of accelerators that are going for throughput." "And they go, 'Yeah, but we're not going to be able to compete with the other guys at high interactivity.'" "And OpenAI has a chip that can do both for them, really well." "I mean, at a minimum, this thing is doing really well right now and is going to serve real production tokens for them." _______ More takeaways from OpenAI's Jalapeno development: firesidealpha.substack.com/p…
149
174
211
2,874
4,572,363