Home page
Quantum AI is useless.
(Fig.1) Quantum mechanics or computers cannot simulate materials, contrary to hypes.

The recent overhyped news falsely claimed that IBM (= too error-prone to give right answers ) could simulate some real magnetic material, and would potentially outperform a classical computer.
This or this site admits No quantum computer advantage nor utility ↓
Key insights says -- No quantum computer advantage
"Classical computers are faster and more accurate than the quantum variety."
"A quantum advantage is impossible to reach with current hardware"
4th-paragraph says
"The researchers focused on crystalline KCuF3, a well-studied magnetic material. They simulated neutron scattering via quantum calculations (= irrelevant to quantum mechanical prediction ) and compared the results with data"
6th-paragraph says -- Classical computer better
"Furthermore, the classical simulation conducted in this study beat the quantum version on every metric the researchers tested in terms of accuracy and computing efficiency."
This-6th-last-paragraph says -- Useless simulation
"The researchers emphasize that the result does not eliminate the need for more advanced hardware. The simulations were performed under constraints that required careful optimization (= meaning No quantum mechanical prediction ), and scaling the approach to more complex materials will require improvements in both qubit quality and system size"
↑ This IBM research paper (= this-2nd-paragraph link ). ↓
p.2-left-1st-paragraph says -- Unreal spinon quasiparticle
"We perform a direct experimental validation by benchmarking quantum simulations
against INS (= inelastic neutron scattering ) measurements of the quasi-1D antiferromagnet KCuF3, whose spectrum is a canonical realization
of.. spinon (= unreal quasiparticle model split from an electron ) . Using 50-qubit (= still Not a quantum computer ).."
p.3-Fig.2(A) says -- Unreal spinon quasiparticles scatter neutrons ?
" Schematic of the spread of fractionalized excitation ( spinon = unreal old quasiparticle with nagative mass )
following neutron scattering"
↑ The same research paper ↓
p.3-(3) showed -- Fake model, free parameters
only abstract unphysical old Hamiltonian (= total energy ) XXZ equations (= lacking real particle picture ) allegedly expressing these magnetic materials as fictional spinon quasiparticles with freely-adjustable parameters J and ε irrelevant to quantum mechanical prediction ( this-p.10-5 ).
p.6-right-1st-paragraph -- Choose parameters, No quantum prediction
artificially chose free parameters J and ε (= adjusted based on experiments like INS, Not on quantum mechanical prediction, this-p.5-left-last-paragraph ) of this unphysical spin model.
p.11(or S3 )-last-paragraph says -- Adjust free parameters for qubits
"we employ approximate quantum compilation (AQC) based on the tensor network approach (= this AQC just freely adjusts parameters for manipulating qubits to produce desired results with No quantum mechanical prediction, this-p.5-3-Algorithm )."
p.11-last-paragraph says -- Error-prone quantum computer
"We optimized the quantum circuit over a short time period
and achieved circuit fidelity of 90 % (= error-rate 10% ) compared to the time-evolved state obtained from matrix-product-state (MPS)
simulation (= MPS is a noiseless = errorless classical computer, this-p.1-left-2nd-paragraph )"
p.12-2nd-paragraph says -- Classical computer needed
"To prepare the ground-state circuits for the spin models studied.. The optimization is performed classically (= an ordinary classical computer was needed to optimize quantum qubits ), with
the loss function chosen"
p.17(or S9)-J says -- Better errorless classical computer
"protocols using MPS simulations (= run classical computers, this-p.1-left-last-paragraph ), which serve as noiseless benchmarks (= MPS run on classical computers is an ideal noiseless = errorless classical simulator outperforming the error-prone or noisy quantum computer, this-p.5-right-last-paragraph )"
This research just tried to express some magnetic material such as KCuF3 as the fictional spinon quasiparticle model with freely-adjustable parameters fitted to experiments like INS (= inelastic neutron scattering, this-p.5-left-last-paragraph ) with No quantum mechanical prediction .
Parameters for flipping qubits (= of the useless IBM quantum computer ) were also artificially chosen to get the desired results in AQC approximation ( this-p.17-J ) with No quantum mechanical prediction ↓
This-p.5-3-Algorithm says -- Free parameters, No quantum prediction
"Approximate quantum compiling (AQC) involves the design of a parametric quantum circuit with fixed
depth - the parameters are then adjusted to bring it as close as possible to the target (= No quantum mechanical nor computing prediction )"
Quantum mechanics, which cannot solve its Schrodinger equation nor predict anything, has to choose fake trial wavefunctions in its approximate variational (= Hartree-Fock ) methods and one-single-fake-electron DFT model that are too time-consuming, useless, cannot predict anything ↓
This-p.6-left used this incorrect quantum mechanical approximation = density functional theory (= DFT ) treating the whole (magnetic) material as one fake electron, which can Not predict any physical values due to its artificially-chosen pseudo-potentials called exchange-correlation energy functionals (= exact form is unknown ) and Coulomb, exchange parameters U,J ( this-p.1 )
Today's error-prone (= noisy ) quantum computers were inferior to the errorless (= noiseless ) classical computer simulation called MPS.
IBM is at the center of constantly leading the world into a rabbit hole of the current fake useless mainstream quantum physics by exploiting its leading quantum computer technology (= Qiskit = impractical forever ) and the mainsream science power (= topics conducted by IBM's world latest quantum computers are easily accepted by top journals and treated as mainstream physics ), cooperating with academia.
The current mainstream quantum mechanics based on fictional parallel worlds is useless, cannot solve its Schrodinger equation nor predict any multi-electron atomic energies, much less discover drugs, contrary to a lot of hypes.
So quantum computers (= scam ) can Not change the fact that quantum mechanics is impractical, cannot predict any atomic or molecular energies in deceptive hybrid quantum (= just classical ) computing methods called variational quantum eigensolver (= VQE ), SQD in seemingly simulating proteins ( this-lower ) mainly led by IBM-academia circles, and Shor algorithm that cannot factor even 21.

Feel free to link to this site.