This builds on the Kickmix quantum language from Google. Essentially, Kickmix is an assembly language for classically simulatable quantum circuits, and finding more efficient circuits is basically a game of code golf with the assembly instructions. That means LLMs benefit from the closed loop of verification, and can rapidly test and discard ideas. It's not too surprising that previous academics didn't try to optimize to this level, and a good task to automate. See ecdsa.fail for another example of circuit optimization using the Kickmix language.
It's a nice illustration of the Envelope Principle: AI is good at filling in the blanks, where the knowledge is 'within the envelope' of things already discovered but the area itself is still not fully exhausted. Real progress happens by expanding the area of the envelope, not by coloring in the blanks.
> Real progress happens by expanding the area of the envelope, not by coloring in the blanks.<p>Respectfully disagree. What one person may classify as expanding the envelope, another one with broader knowledge may classify as coloring in the blanks. Much knowledge in one field comes from cross-pollination from other fields.<p>Let me give you an example I just learned about. The database community spent decades trying (to some extent) to solve the implementation of materialized views to arbitrary complex SQL queries. Then, a few years ago, someone with experience in digital signal processing brought their expertise to the DB field and completed the puzzle. [1]<p>Someone with DB expertise but no signal processing expertise might say that is expanding the envelope, but another person with experience in both may say it is coloring in the blanks.<p>AI work now may seem like coloring in the blanks, but given it's vast expertise in multiple fields, my bet is that for most people it will look like pushing the envelope very soon.<p>[1] <a href="https://arxiv.org/abs/2203.16684" rel="nofollow">https://arxiv.org/abs/2203.16684</a>
People (most people?) have spent their entire decades long careers coloring in the blanks within the envelope of human knowledge. I think it’s kind of shitty to suggest that they weren’t making “real” progress with their lives, just because AI can now do the same thing.
Of course it is progress. But there are orders here: filling in the blanks is super useful and for sure you can - or should I say could - make a living with that. But there is another level above that which is where the total area is increased. That's not meant as a sleight against people that were ensuring all the ground was covered, it's just establishing the fact that there are multiple levels to knowledge.<p>If you want to make it seem as if I'm pissing on people that were doing useful science that's your privilege, but there is a reason we use terms like 'groundbreaking' about discoveries that are different.<p>As for lost careers: I've been schlepping bits my whole life, and probably an RTX 6000 could re-run that in under a year, and no doubt what I did was of less significance than that done by the lowliest of scientists. Still, I don't derive my sense of self value from what future technology can do any more than I think that a day laborer should compare themselves with a tractor or a combine.
I'm not sure of your background, but FWIW everyone in my PhD program in my elite CS program had varying levels of anguish and angst when realizing that their efforts were mostly suited for "coloring in the blanks." The PhD Grind [1] felt like a poignant capture of those feelings.<p>[1]: <a href="https://dokumen.pub/the-phd-grind-a-phd-student-memoir-3.html" rel="nofollow">https://dokumen.pub/the-phd-grind-a-phd-student-memoir-3.htm...</a>
>> Real progress happens by expanding the area of the envelope, not by coloring in the blanks.<p>This is a pretty shitty thing to say, and if you think about it for even a few minutes, you'll understand why.
[flagged]
> In other cases, it found that existing circuits included costly components that were needed to handle only extremely rare inputs, and it redesigned them without those components.<p>So it’s not performing the same computation…
This is definitely a level of universal paper clips we’re on…<p>This is the level where the AI develops quantum computing to make itself smarter (in the game)
I'm normally more of an AI hater but this just seems objectively good so I'll ignore it and move on
TLDRAIBBQ: Claude redesigned eight simulated quantum circuits, reducing operations by up to 17× and 12× by reconstructing temporary data and replacing an algorithm with a better one.
Liar Claude:<p>"This work was led by Andi Gu as part of Anthropic’s STEM Fellows Program, with mentorship from Alek Kemeny, a quantum physicist at Anthropic, in collaboration with Susanne Yelin (Harvard University) and Mikhail Lukin (Harvard University and QuEra Computing). Read the full paper on arXiv."<p>The actual paper (<a href="https://arxiv.org/pdf/2610.22367" rel="nofollow">https://arxiv.org/pdf/2610.22367</a>):<p>"We thank Pablo Bonilla, Arjun Kar, Nazli Ugur
Koyluoglu , Rohan Mehta, Varun Menon, and Vinay Ra-
masesh for feedback on the manuscript. AG conducted
<i>part of this work</i> under the Anthropic STEM Fellows
Program. Work at Harvard was supported by IARPA
and the Army Research Office under the Entangled Log-
ical Qubits program (Cooperative Agreement Number
W911NF-23-2-0219) and by the National Science Foun-
dation (NVQL grant PHY-2410716)."<p>The whole Anthropic commercial appears to be hallucinated.