Physicist by training silly by choice. Views my own. Account https://xitter.gq/t.co/L8zG143Woe at https://xitter.gq/t.co/ZGNQfE6hRu
Canada
Joined April 2013
- Tweets71.4K
- Following2.2K
- Followers4K
- Likes31.2K
Marco Piani 🇮🇹🇨🇦 retweeted
Ah, another beautiful question in quantum complexity theory solved! This one is due to Scott Aaronson and @AAmbainis, about the necessity of structure in a problem for exponential quantum advantage. This question inspired a lot of research and we collectively learned a lot about quantum computing from it.
The problem seemed ripe; this year it felt like we were closing in on it. The most recent (human-written) papers on it had solved special cases showing that parallel quantum algorithms cannot achieve exponential speedups for unstructured problems.
What is the next beautiful question in quantum algorithms that will spur another generation of learning and understanding?
* Guy Blanc, Jordan Docter, Carmen Strassle and Li-Yang Tan, Quantum Speedups Require Structure or Depth (arxiv.org/abs/2608.19158)
* Qipeng Liu, Saachi Mutreja (my fantastic student!), Parallel Quantum Advantage with Limited Adaptivity Requires Structure (arxiv.org/abs/2608.20297)
Today Physical Superintelligence PBC is announcing a proof of the Aaronson-Ambainis conjecture, one of the central open problems in quantum complexity, with major implications for quantum computing speedup.
psi.inc/research/aaronson-am…
Thread 👇
It is somewhat interesting to see Tesla fans/owners/shareholders complain about the choices Tesla/Musk makes.
Musk is not interested in selling cars. He has these overarching/enormous goals, like ensuring the future/survival of humanity/consciouness, or spreading the latter…
This was inevitable at the levels it was selling at. However, it continues a disturbing trend. S/X were killed for Optimus production. Cybertruck dies for Cybercab production. What thing that most Tesla owners don't give a crap about will kill 3/Y production?
Tesla is trending toward no longer being a car company. It depresses me. 😞
…to the stars, or understanding the Universe.
Anything he does is motivated at that level.
So, if he thinks that what is best for humanity is to have fleets of autonomous cars, autonomous robots, etc, and that that is the future, he will not care about building “standard” cars
$TSLA is currently trading at a bit more than one third of the *bear* case for 2026 of the @ARKInvest valuation model of years ago.
“What went wrong?”, you could ask.
A potential answer: real life happened.
What could @Tesla be worth in 2031? ARK’s long-awaited valuation just dropped.
ark-invest.com/articles/valu…
.@elonmusk should not self-censor in this way after spending tens of billions to ensure freedom of speech!
Marco Piani 🇮🇹🇨🇦 retweeted
I’m struggling to catch up with the capabilities of models I already have access to.
Today I gave Opus free rein to improve an old transformer project I slopped together earlier this year. I thought the results were stunning.
github.com/tobiasosborne/Pic…
The best thing about this kind of takes is that they sound so certain about what would have needed to happen for the result to be the official one and lawfully achieved.
E.g., he could have lost some of the votes he received in the first round, but got even more new votes.
Para Lula chegar aos 64,2 milhões de votos, ele teria que ter:
- Mantido 100% dos votos que recebeu no primeiro turno
- Conquistado 100% dos 3 milhões de eleitores que deixaram de se abster
- Recebido 100% dos 1,8 milhão de votos que antes foram anulados
- Conquistado cerca de 60% dos votos dos outros candidatos (Augusto Cury, Renan Santos, Caiado e Zema)
By "lost", I mean that some people could have switched sides, or might not have voted in the second round even if they had voted in the first one.
If you only know the subtotals, there is a lot that could have lawfully happened, still leading to this result.
Marco Piani 🇮🇹🇨🇦 retweeted
Quantum entanglement can give precision sensing a major edge, finds a new study. Researchers show quantum speedups for learning classical signals, opening new possibilities for electromagnetic sensing and gravitational wave detection.
📄 go.aps.org/4iaoUKl
ALT A two-panel diagram comparing conventional and quantum-enhanced sensing of classical fields. In the top panel, conventional sensing measures a complex field along selected axes, producing separate projections that lose information, or a blurred distribution due to added vacuum noise. In the bottom panel, quantum-enhanced Bell Quantum Signal Learning uses entangled two-mode squeezed states, with one mode probing the field and the other stored as a reference. Beam splitters and homodyne measurements combine the modes to recover multiple properties of the field with greater precision and without choosing a single measurement direction in advance.
There was a time when AI was seen as a competitor to quantum computing, meaning, for example, that funding/investments in quantum computing could be less because redirected instead to AI.
But one ‘property of AI’ is that it is a general accelerator, …
New paper with @MCrichigno at @Phasecraft. AI agents optimize quantum circuits, and an exact verifier checks every result. arxiv.org/abs/2610.31352
The challenge itself is automated too. An agent works with a researcher to build the verifier, tests it, and freezes it. Then the agents compete.
We observed large gains on every problem we tried, from fermion-to-qubit encodings to qubit routing and unitary synthesis. The method applies to any problem with an exact checker.
It's great to see others working in the same direction this week. Excellent papers from Google Quantum AI and Harvard/QuEra/Anthropic:
arxiv.org/abs/2610.23832
arxiv.org/abs/2610.22367
and the inverse square root challenge:
opensource.googleblog.com/20…