正在加载视频...

视频加载失败

The secret behind quantum search efficiency

3,241,785 次观看 • 7 个月前 •via X (Twitter)

34 条评论

Axel 的头像
Axel7 个月前

Why "quantum" tho? I'd just start a new thread every time I see a new path

Берлинский Олень 的头像
Берлинский Олень7 个月前

It's called BFS and not quantum search 🤡

Dr. Md. Amir Khusru Akhtar 的头像
Dr. Md. Amir Khusru Akhtar7 个月前

It’s not faster because it runs harder — it’s faster because it lets the wrong answers cancel themselves.

Govind 的头像
Govind7 个月前

So grover's algorithm is basically Naruto using Shadow Clone Jutsu to search the whole forest at once, then cancels all clones except the one that found the ramen?

Moravyalı 的头像
Moravyalı7 个月前

That's parallel computing and it's not faster than Dijkstra in terms of computational complexity.

Aanshul Sadaria 的头像
Aanshul Sadaria7 个月前

Why does this look like BFS?

Hawkeye 的头像
Hawkeye7 个月前

So slime mold is a biological quantum search? It searches all possible pathways, and then reinforces only the shortest path branches that have found food.

Rex⚡Volt 的头像
Rex⚡Volt7 个月前

The improbability drive has arrived.

Pat the Irishman 的头像
Pat the Irishman7 个月前

And this is why your binary passwords are worthless against quantum computers

Smit 的头像
Smit7 个月前

I think it'll also help people to understand quantum computing:

TUVegeto137 的头像
TUVegeto1377 个月前

You make it look like quantum computing is just parallel computing. It's not.

𝓛𝓪𝓾𝓭𝓮𝓵𝓲𝓷𝓸 的头像
𝓛𝓪𝓾𝓭𝓮𝓵𝓲𝓷𝓸7 个月前

Seems like a thread

MrBlindMouse 的头像
MrBlindMouse7 个月前

So its just an async Breadth-First search?

Rahul 的头像
Rahul7 个月前

Grover's algorithm is fascinating because it achieves quadratic speedup over classical search - O(√N) vs O(N). But the real challenge isn't the algorithm, it's maintaining quantum coherence long enough to run it at scale. Quantum error correction is the bottleneck we're solving now.

Lukas Süss 的头像
Lukas Süss7 个月前

@sciencegirl — Not true or at least misleading. Quantum comuting is still strictly less powerful compared to true parallel computing due to the random sampling constraint. (Ignoring physical impossibility of such parallelism.)

Δημήτρης Βουζουναράς 的头像
Δημήτρης Βουζουναράς7 个月前

Nothing to do with... quantum anything... The one is a linear pathfinder, while the second is a multi-threaded one, creating a new thread (job, worker) for any possible new path, killing it when it arrives to a dead-end. Quantum computing is still a myth.

Jason Dadness 的头像
Jason Dadness7 个月前

@grok is this a good description of quantum computing?

Hemant 的头像
Hemant7 个月前

So classical computers panic, quantum computers just vibe and arrive.

ItsBS 的头像
ItsBS7 个月前

Quantum Computing is fake. Quantum State Superposition is a lie. Don't be fooled, like so many others, because this idea of "parallelism" has already been proven false in experiments. Classical superposition of the wavefunction exists, but Born's ad-hoc "probabilities" do not.

Craig Swann - model citizen, zero discipline 的头像
Craig Swann - model citizen, zero discipline7 个月前

I've yet to find an explanation of how a bit being a 1 and 0 simultaneously, is useful.

AWeirdRobot 的头像
AWeirdRobot7 个月前

Wrong

Anthony Schmitt 的头像
Anthony Schmitt7 个月前

Stop searching like it's 2020. Classical linear search is officially a bottleneck for Big Data. The future? Quadratic speedup. Grover’s Algorithm isn't just theory—it’s the new strategic edge for 2026.👇 #QuantumComputing #Algorithm #DeepTech @sciencegirl

Miggy Ferrero Rocher 的头像
Miggy Ferrero Rocher7 个月前

Like, oh my gawd, nooo way, that's totally not correct, duh! 🙄

Ludic labs 的头像
Ludic labs7 个月前

The real magic isn’t speed. It’s exploring better paths before committing to one.

MoroAI 的头像
MoroAI7 个月前

Why not look at it from above recognize where the exit is and then take the best path

G Labs 的头像
G Labs7 个月前

This quantum search method for solving mazes looks like a backwards version of the Dead End Filling Algorithm, where false paths are colored in starting from dead ends, leaving the correct path revealed. Here's my mini lecture and demonstration of it:

Elena_ The HumanistPoet 的头像
Elena_ The HumanistPoet7 个月前

I need an agent to navigate my life. Best path forward please.

Ben 的头像
Ben7 个月前

A* algorithm entered the chat. If you're out here solving mazes by trying every goddamn dead-end like a drunk toddler with a death wish... congrats, you've discovered the slowest, most soul-crushing way possible. Real maze algorithms don't wander around like lost tourists.

Steve Nicklin 的头像
Steve Nicklin7 个月前

I might be wrong but is the same approach that CHIP (Constraint Handling in Prolog) took in the 1980's (yes you read that right 1980's!) Try Quantum Search against the 8 Queens Problem on a 64x64 chess board and see if it is as quick!

Göktuğ 的头像
Göktuğ7 个月前

This algorithm is not entirely suitable for parallelization. Some agents may have to follow the path for a very long time, which actually makes the algorithm inefficient to run both on CPU and GPU.

Hold this L~ 的头像
Hold this L~7 个月前

if quantum can search at multiple path in the same time, why cant normal search just go multiple path search too?

SereneLoop 的头像
SereneLoop7 个月前

Wow thats great fo you that you know about quantum search efficiency

Potato 的头像
Potato7 个月前

It's called BFS, and it's computationally heady in space. You're trading memory for time, as always. There is no miracle solutions.

Cloud. 的头像
Cloud.7 个月前

And that's exactly how we humans are built in multiple universes.

相关视频