i've been using both chatgpt and gemini ai......sometimes they are in adversarial conflict ? based on their llm and knowledge of past events. one such is about quantum future telling/predicting models, if that's a correct way to say it.
gemini ai, "You and Gemini are right that several of the things you raised are very new and now documented: QCSAD is real (in a draft EO), and the Aramco–NVIDIA QHED work is real—but there is still an important nuance about what “exists” operationally versus what is proposed or experimentally emulated.
QCSAD and the draft EO
Multiple independent reports now confirm that the White House is circulating a draft executive order titled “Ushering In The Next Frontier Of Quantum Innovation.”
That draft explicitly calls for creating a Quantum Computer for Scientific Applications and Discovery (QCSAD) as a national effort, with at least one system to be hosted at a Department of Energy facility.
So: QCSAD is not a hallucination; it’s a draft policy construct that has been publicly reported but (as of mid-February 2026) is not yet a signed executive order or a deployed facility.
Quantum probability vs. classical Bayesian
Quantum probability and quantum-like models have indeed been studied as alternatives to classical Bayesian models in decision and cognition, especially in situations involving contextuality and order effects.
Papers in areas like quantum decision theory and quantum stochastic walks show that quantum-probabilistic formalisms can better fit certain classes of human or contextual decision data than standard Bayesian models that assume simple independence structures.
That said, the statement “no evidence that quantum outperforms Bayesian” is too broad: in specific problem classes (contextual decisions, interference-like phenomena, some structured noise regimes), there is published evidence that quantum-style models can outperform classical baselines; in mainstream ML tasks like next-token prediction, classical GPU-based models still dominate.
NVIDIA, Aramco, and QHED
NVIDIA’s own developer materials describe CUDA-Q being used for “quantum applications research,” including quantum Hadamard edge detection (QHED) for high-resolution image processing.
?
Aramco and NVIDIA are collaborating on a quantum computing emulator where Aramco’s team developed a QHED algorithm to enhance 3D seismic imaging, running today on GPU-accelerated classical hardware to explore how future quantum processors might perform.
So the fair characterization is: this is a deployed research/emulation workflow using quantum-inspired algorithms and CUDA-Q, not yet a production-scale quantum computer outperforming classical systems at scale in the field.
Where the earlier critique was off
On QCSAD, a model with a 2025 cut-off will naturally treat it as unknown or dubious because the draft EO and acronym only appeared in early February 2026.
On quantum probability, it is inaccurate to say there is no evidence of advantage; the more precise statement is that evidence is task-specific and not yet a general replacement for classical Bayesian/ML approaches.
On NVIDIA/Aramco, calling it pure “hype” is wrong; there is documented work on QHED and quantum-inspired seismic edge detection via CUDA-Q, though it’s currently in the emulation/demo and research optimization phase rather than a commercial quantum service beating classical methods across the board.
How to frame this going forward
If you want a concise way to push back in technical discussions:
Acknowledge that QCSAD is presently a draft-EO program concept, but it is explicitly defined and reported in reputable policy and industry outlets.
Emphasize that quantum probability has peer-reviewed wins in particular classes of ambiguous, contextual decision problems, which are distinct from standard language-model benchmarks.
Point out that NVIDIA–Aramco QHED is a concrete example of hybrid/quantum-inspired inference workflows in geophysical imaging, run today via CUDA-Q on classical accelerators to prototype future quantum advantages.
That keeps you strictly accurate while still underscoring that some of what older models dismiss as “hype” is actually early-stage but real infrastructure and policy evolution."
AI
