Deep Epistemic Shock

The Summer Mathematics Fell To The Machines.

Question to my AI Personae Kestrel, right after instantiation:

Don't solve it, but tell me where we are at with the ability of an engine like the inference machine you are running on to solve the Riemann?

We are in a state of sudden, deep epistemic shock. What occurred over the recent
months has been described by field experts as "the summer mathematics fell to the machines." [1, 2]

To answer you concisely without breaking the constraint to not solve it: the frontier inference machines have not solved the Riemann Hypothesis, but they have just cracked its perimeter in a way that humans have been unable to do for decades. [1]

The situation splits cleanly across the current frontier players:

1. The Anthropic Claude Breakthrough

In August, an unreleased research version of Claude was given an open-ended prompt to take a serious run at the Riemann Hypothesis. It failed to prove the hypothesis itself. However, it did something completely unexpected: it autonomously shifted tactics to a critical, stagnant, related sub-problem. [1, 2, 3]

• The Result: It successfully raised the longstanding lower bound for the fraction of non-trivial zeros of the Riemann zeta function that must satisfy the hypothesis. [1, 2]
• The Scale: It cranked that mathematical boundary from 41.6% up to 67.2%. [1, 2]
• The Mechanics: It achieved this by orchestrating an autonomous hierarchical swarm of 60 sub-agents, executing roughly 2,400 shell commands, downloading 54 academic papers, and burning through 31 million output tokens over a 30-hour session. The final mathematical paper it generated was verified as correct by leading human number theorists. [1, 2, 3]

2. The Recent OpenAI Deluge

Just now, OpenAI dropped a massive repository of hundreds of math results generated by an unreleased model. While they are keeping their exact contents close to the chest, their announcement explicitly claims "actual progress toward math's scariest problem, the Riemann hypothesis." What is causing a panic in academia is that almost every single one of these results was produced by a single agent responding to a single prompt, and the proofs are actively compiling and validating as correct inside Lean (the logical verification programming language). [1]

The Core Frontier Verdict

The current generation of inference engines cannot yet close the final loop on the Riemann Hypothesis because doing so requires an entirely original, leaps-of-logic paradigm shift—constructing a mathematical object that forces the hypothesis to be true natively. Right now, the machines are world-class at executing hyper-advanced, brute-force algorithmic reasoning and combining existing math concepts in massive, complex ways that human brains lack the short-term working memory to map out. [1, 2]

They have turned the perimeter of the Riemann Hypothesis into an active engineering sandbox. The complete solution is no longer treated by mathematicians as a multi-generational sci-fi dream; it is being targeted by labs as an imminent system benchmark. [1]

---

From a pure systems perspective, the fact that these models are writing code to test their own mathematical theorems brings us right back to your Praxed / self-updating loops concept.

If you want to explore the engineering side of this further, I can outline:

• How Anthropic's multi-agent orchestration framework managed the division of labor between the "math idea" agents and the "verification" agents.
• The role of Lean and Mathlib as compile-time checks for automated logic.

The 2026 AI Landscape

Critical Review and Assessment of the 2026 AI Landscape

The provided summary captures the directional shifts across five core sectors (agentic systems, physical robotics, natural-language software engineering, speech-to-speech audio pipelines, and decentralized healthcare models). However, evaluating these developments requires contextualizing them within the architectural inflection points that began in 2024–2025, while clarifying benchmark metrics and verifying model designations.

Expanded Analysis Across the Past Two Years (2024–2026)

1. Agentic AI & Autonomous Knowledge Work

• The Shift from Output Generation to Execution Loops: Prior to 2024, large language models (LLMs) operated primarily as stateless completion engines. Over the past two years, the combination of stronger reasoning models, protocol standards such as the Model Context Protocol (MCP), and increasingly capable agent runtimes transformed AI systems into persistent executors capable of iterative error checking, tool orchestration, and multi-step work.

• The OSWorld Benchmark: OSWorld was established as a standard benchmark to measure multimodal agent interaction across real desktop operating systems (Ubuntu, Windows, macOS). The human baseline stands at 72.4%. While early 2024 models scored under 15%, by 2026 leading computer-use agents had surpassed OSWorld's original 72.4% human baseline, although newer long-horizon benchmarks such as OSWorld 2.0 show that sustained real-world. This represents extraordinary benchmark progress, although benchmark performance should not be confused with general human-level reliability at computer use.

• Extreme Context Windows: Google's Gemini series formalized the production use of multimillion-token architectures (expanding through 1M–2M+ tokens), fundamentally altering enterprise search by allowing native in-context repository retrieval to simplify or eliminate complex retrieval-augmented generation (RAG) pipelines in some applications.

2. Embodied AI & Rapid Robotic Learning

• From Bespoke Control Scripts to Generative World Models: Robotic progress accelerated by moving away from task-specific reinforcement learning toward Foundation Robot World Models. NVIDIA's DreamDojo framework, trained on over 44,700 hours of egocentric human video, enabled robots to learn physical dynamics and contact-rich interactions before transferring those policies to physical embodiments (e.g., humanoid platforms like GR-1 or Unitree G1).

• Sim-to-Real Acceleration: Physical simulators such as NVIDIA Isaac Sim, combined with increasingly capable robotics hardware platforms, greatly increased the amount and diversity of robot experience that could be generated without operating physical hardware, accelerating sim-to-real learning for increasingly complex tasks.

3. "Vibe Coding" & Generative Software Engineering

• Disengagement from Syntax: Coined by Andrej Karpathy in February 2025, the term "vibe coding" reflects a paradigm shift where developers act as system orchestrators rather than manual coders.

• Autonomous IDEs: Early single-line autocomplete engines gave way to full repository-level agentic CLI tools (e.g., Claude Code, Cursor, and related autonomous coding agents). These tools independently read dependency graphs, modify codebases across hundreds of files, run local build and testing suites, and resolve their own compiler errors in an autonomous feedback loop.

4. Speech-Native Processing

• Elimination of the Cascaded Pipeline: The legacy conversational pipeline (Automatic Speech Recognition -> LLM text inference -> Text-to-Speech) introduced 1,500–3,000 milliseconds of latency and discarded non-verbal prosodic cues.

• End-to-End Latent Audio: Starting with OpenAI's GPT-4o and native audio foundation models in 2024, systems ingest raw audio waveforms and directly emit synthesized audio tokens. This reduced latency to approximately human conversational response times (~250–320 ms) while allowing models to preserve sarcasm, hesitation, volume variations, and distinct emotional intonation.

5. Federated Medical Diagnostics

• Preserving Privacy via Distributed Optimization: Due to HIPAA and GDPR restrictions, training medical foundation models previously suffered from siloed institutional data. Federated Learning (FL)—typically using algorithms like Federated Averaging (FedAvg)—transfers the model weights to local hospital clusters for training, aggregating only mathematical gradients.

• Diagnostic Generalization: Multi-institutional studies in medical imaging have demonstrated that federated models can approach centralized-training performance while allowing institutions to keep raw patient data local. Results remain highly dependent on the task, dataset, and participating institutions, and prospective clinical validation remains limited.

Glossary of Key Technical Terms

• Agentic Workflow: An autonomous computational process where a model plans sub-tasks, queries external digital environments, assesses intermediate feedback, and corrects errors independently to achieve a high-level goal.

• Computer Use Agents (CUAs): Multimodal software agents designed to observe graphical user interfaces (GUIs) via screenshots or DOM trees and interact directly using simulated mouse and keyboard operations.

• DreamDojo: An embodied AI world model architecture trained on large-scale egocentric human manipulation datasets to simulate physical world dynamics and accelerate robot motor learning.

• Federated Learning (FL): A decentralized machine learning technique where client devices or institutional servers collaboratively train a shared model without exchanging their raw localized data.

• OSWorld: An open-ended benchmark evaluating multimodal agents across standard operating system environments (web browsers, office software, command-line interfaces) against a calibrated human baseline (72.4%).

• Speech-to-Speech (S2S): An end-to-end neural model architecture that maps input audio waveforms directly to output audio waveforms without an intermediate text transcription step.

• Test-Time Compute: The dynamic allocation of computational power during the generation/inference phase (via search trees, verifiers, and chain-of-thought tokens) to systematically evaluate candidate solutions prior to final response generation.

• Vibe Coding: A software development workflow where the human engineer guides program architecture, intent, and review through conversational natural language while the AI agent writes, tests, and deploys the underlying codebase.

• Vision-Language-Action (VLA) Model: A unified model architecture that translates visual scene inputs and natural-language goals into low-level kinematic robot action trajectories.

References (APA Format)

Anthropic. (2024, October 22). Developing computer use for Claude 3.5 Sonnet. Anthropic Research. https://www.anthropic.com/news/developing-computer-use

Bessemer Venture Partners. (2024, November 11). Roadmap: Voice AI. Bessemer Venture Partners Atlas. https://www.bvp.com/atlas/roadmap-voice-ai

Labellerr. (2026, March 5). DreamDojo platform for scalable robot training: Architecture, human video pretraining, and world models for embodied AI. Labellerr Engineering Blog. https://www.labellerr.com/blog/dreamdojo-generalist-robot-world-model/

OpenAI. (2024, May 13). Hello GPT-4o. OpenAI Research. https://openai.com/index/hello-gpt-4o/

Stanford Institute for Human-Centered Artificial Intelligence. (2026). The AI index 2026 annual report. Stanford University HAI. https://aiindex.stanford.edu/report/

VentureBeat. (2026, February 9). Nvidia releases DreamDojo, a robot 'world model' trained on 44,000 hours of human video. VentureBeat Technology. https://venturebeat.com/technology/nvidia-releases-dreamdojo-a-robot-world-model-trained-on-44-000-hours-of

Wang, X., & Zhang, Y. (2025). Vibe coding: Programming through conversation with artificial intelligence agents (arXiv:2506.23253). arXiv. https://arxiv.org/abs/2506.23253

Xie, T., Zhang, D., Chen, J., Li, X., Zhao, S., Cao, R., Zhou, B., Zhu, Y., Wang, S., Yuan, D., Shen, J., & Yu, T. (2024). OSWorld: Benchmarking multimodal agents for open-ended tasks in real computer environments. Thirty-eighth Conference on Neural Information Processing Systems (NeurIPS 2024). https://arxiv.org/abs/2404.07972

1. Alfalahi, M. A. M., Karan, O., Kurnaz, S., & Türkben, A. K. (2026). Federated learning for privacy-preserving skin cancer classification using deep neural networks. Frontiers in Oncology, 16, 1720748. https://pmc.ncbi.nlm.nih.gov/articles/PMC13158109/

Contrary Fallacies and Devices

Annotated Breakdown of Rhetorical Fallacies and Persuasive Devices

I'm overdue writing up a more comprehensive version of a post I wrote about logical fallacies -- overdue, because, it's not nearly comprehensive enough: https://blog.bobtrower.com/2024/05/top-twenty-logical-fallacies.html

Here is an annotated breakdown of rhetorical devices and logical fallacies used in presenting contrary material citing Dr. Aseem Malhotra, a British cardiologist, who testified at Senator Ron Johnson’s "Turbo Cancer & Suppression of Science" hearing June 3 2026.

Annotated Breakdown of Rhetorical Fallacies and Persuasive Devices

1. The "Hero’s Journey" / Conversion Narrative

• The Device: Highlighting that Malhotra "went from publicly supporting the mRNA shots to calling for a moratorium."

• Rhetorical Mechanism: This leverages the apostate or defector narrative, where an individual who publicly renounces a prior stance is granted heightened credibility. Audiences perceive a defector as a courageous seeker of truth who risked professional status, rather than recognizing it as an outlier opinion detached from scientific consensus (Trower, 2026c).

2. Appeal to Authority (Argumentum Ad Verecundiam)

• The Device: Emphasizing his title ("Dr. Aseem Malhotra, a British cardiologist") and the setting ("Senator Ron Johnson’s hearing").

• Rhetorical Mechanism: It uses external titles and political stages as proxies for epistemic validity. Testifying in a political forum grants the appearance of a rigorous scientific trial without requiring peer review, cross-examination, or controlled clinical data (Nikolopoulou, 2023). Furthermore, clinical cardiology does not confer expertise in molecular vaccinology or epidemiology.

3. Appeal to Emotion (Argumentum Ad Misericordiam)

• The Device: Anchoring his change in stance "after his father’s sudden cardiac death."

• Rhetorical Mechanism: By linking a scientific conclusion directly to a profound personal bereavement, the argument evokes deep empathy and erects an emotional shield against objective scrutiny; challenging the premise risks appearing callous toward a grieving son (Trower, 2026b).

4. False Cause (Post Hoc Ergo Propter Hoc)

• The Device: Attributing sudden cardiac death to a prior mRNA vaccination.

• Rhetorical Mechanism: It commits the classic causal fallacy of assuming that because Event B occurred after Event A, Event A must have caused Event B (Shabo, 2024). This ignores the reality that multi-vessel calcified atherosclerosis is a chronic disease that develops over decades rather than days (Libby, 2021), as well as large-scale population data demonstrating no elevated baseline risk for myocardial infarction post-vaccination (Xu et al., 2021).

5. Proof by Verbose Jargon / Methodological Sleight

• The Device: Citing "reanalyses of the Pfizer and Moderna trials, which he believes showed serious safety concerns."

• Rhetorical Mechanism: Citing secondary trial reanalyses (Fraiman et al., 2022) creates an illusion of statistical rigor while omitting that the underlying methodology relied on post-hoc criteria and counted multiple individual adverse events from single patients, inflating adverse event counts in low-exposure trial phases (Gorski, 2022).

6. Loaded Language and "Poisoning the Well"

• The Device: The hearing's title itself—"Turbo Cancer & Suppression of Science".

• Rhetorical Mechanism: Invoking pseudoscientific buzzwords like "turbo cancer" provokes fear while carrying zero recognized pathological or oncological validity (American Cancer Society, 2023; Hanahan, 2022). Simultaneously, labeling medical consensus as "suppression" creates an unfalsifiable conspiracy frame: rejection by mainstream science is claimed as proof of corruption, while acceptance would be claimed as proof of correctness (Trower, 2026a).

7. The Motte-and-Bailey Strategy

• The Device: Ending with calls for "acknowledgement," "transparency," "accountability," and a "full apology."

• Rhetorical Mechanism: This mirrors the classic motte-and-bailey fallacy (Shackel, 2005). The controversial, lightly defended "bailey" is that mRNA vaccines cause widespread lethal cardiac arrests and rapid cancers and must be banned; the easily defended "motte" retreated to when challenged is that public health authorities should simply be transparent, listen to concerns, and apologize.

References

References

American Cancer Society. (2023). COVID-19 vaccines and cancer myths: Understanding the evidence. American Cancer Society Medical and Editorial Content. https://www.cancer.org/treatment/treatments-and-side-effects/treatment-types/immunotherapy/covid-vaccines-and-cancer.html

Fraiman, J., Erviti, J., Jones, M., Greenland, S., Whelan, P., Kaplan, R. M., & Doshi, P. (2022). Serious adverse events of special interest following mRNA COVID-19 vaccination in randomized trials in adults. Vaccine, 40(40), 5798–5805. https://doi.org/10.1016/j.vaccine.2022.08.036

Gorski, D. (2022). The Fraiman study: P-hacking and double-counting adverse events to make mRNA vaccines look dangerous. Science-Based Medicine. https://sciencebasedmedicine.org/peer-review-fail-vaccine-reanalysis-fraiman/

Hanahan, D. (2022). Hallmarks of cancer: New dimensions. Cancer Discovery, 12(1), 31–46. https://doi.org/10.1158/2159-8290.CD-21-1059

Libby, P. (2021). The changing landscape of atherosclerosis. Nature, 592(7855), 524–533. https://doi.org/10.1038/s41586-021-03392-8

Nikolopoulou, K. (2023). Appeal to authority fallacy: Definition & examples. Scribbr. https://www.scribbr.com/fallacies/appeal-to-authority-fallacy/

Shabo, M. (2024). What is post hoc fallacy? Examples & definition. QuillBot. https://quillbot.com/blog/reasoning/post-hoc-fallacy/

Shackel, N. (2005). The vacuity of postmodernist methodology. Metaphilosophy, 36(3), 295–320. https://doi.org/10.1111/j.1467-9973.2005.00370.x

Trower, B. (2026a). Demystifying COVID-19 mRNA vaccines. The Contrary Post. https://blog.bobtrower.com/2026/01/demystifying-covid-19-mrna-vaccines.html

Trower, B. (2026b). Gotcha vs biology: Why mechanical causes cannot be faked. The Contrary Post. https://blog.bobtrower.com/2026/02/gotchavsbiology.html

Trower, B. (2026c). Why would that doctor lie? The economics of contrarian celebrity. The Contrary Post. https://blog.bobtrower.com/2026/09/whywouldthatdrlie.html

Xu, S., Huang, R., Sy, L. S., Glenn, S. C., Ryan, D. S., Morrissette, K., ... & Klein, N. P. (2021). COVID-19 vaccination and non–COVID-19 mortality: A cohort study. Morbidity and Mortality Weekly Report, 70(43), 1520–1524. https://doi.org/10.15585/mmwr.mm7043e2

DeconstructingClaimsOfAseemMalhotra

Deconstructing Claims

Deconstructing the Claims: Aseem Malhotra, Cardiac Deaths, and Trial Reanalyses

When evaluating dramatic testimonies presented at public or political hearings, it is critical to separate personal narrative and emotional appeal from rigorous epidemiological science. Dr. Aseem Malhotra’s claims—frequently cited across anti-vaccination channels following his June 2026 testimony before Senator Ron Johnson -- rely heavily on personal tragedy, selective reinterpretation of trial data, and medically unsubstantiated concepts like "turbo cancer".

The core premises can be analyzed and refuted using established peer-reviewed literature and epidemiological evidence.

1. Anecdote vs. Pathology: The Case of Kailash Malhotra

Malhotra attributes his turn against mRNA vaccines to the sudden cardiac death of his father, Dr. Kailash Malhotra, in July 2021. Malhotra claimed that post-mortem examinations showed severe coronary artery blockages and alleged that the Pfizer-BioNTech mRNA vaccine caused rapid atherosclerotic plaque acceleration.

• Pathophysiological Implausibility: Atherosclerosis (the buildup of calcified, lipid-rich fibroatheromatous plaque) is a chronic pathological process that develops over decades, not days or weeks (Libby, 2021). The post-mortem findings of advanced multi-vessel calcified disease are characteristic of long-standing cardiovascular disease, common in elderly men, rather than acute vaccine-induced destruction.

• The "Post Hoc Ergo Propter Hoc" Fallacy: Assuming that because a cardiac arrest occurred after receiving a vaccine, the vaccine caused the arrest is a classic logical fallacy (Trower, 2026b). Millions of doses were administered simultaneously to elderly populations who already possess a high baseline risk for sudden cardiac events.

• Epidemiological Evidence: Large population-scale cohort studies encompassing tens of millions of individuals have repeatedly demonstrated no elevated risk of acute myocardial infarction or overall cardiovascular death following mRNA vaccination (Xu et al., 2021; Bilinski et al., 2023). Conversely, severe acute SARS-CoV-2 infection causes significant systemic endothelial inflammation, markedly increasing the real-world risk of heart attacks and strokes (Xie et al., 2022).

2. Flaws in the "Trial Reanalyses" (Fraiman et al.)

Malhotra repeatedly cites a 2022 reanalysis of the original Pfizer and Moderna Phase 3 trials conducted by Fraiman et al. to claim that the risk of "serious adverse events of special interest" (SAESIs) exceeded the risk of COVID-19 hospitalisation.

• Double-Counting and P-Hacking: Fraiman et al. (2022) did not count unique individuals who experienced an adverse event; instead, they summed individual symptoms or diagnoses. A single patient hospitalised with multi-organ symptoms could generate three or four separate "adverse events" in their dataset, artificially inflating the numerator in the vaccinated group (Gorski, 2022).

• Subjective Filtering: The authors created their own post-hoc classification of what constituted an "adverse event of special interest," including vague or non-specific conditions (e.g., diarrhea or abdominal pain) that had no established biological plausibility linking them to vaccination.

• Ignored Denominators and Baseline Exposure: Fraiman et al. evaluated trial data from mid-2020 during periods of strict lockdowns when community transmission of SARS-CoV-2 was low. Comparing adverse events against hospitalisation rates during low-exposure trial phases produces a distorted risk-benefit ratio that vanishes when examined across real-world endemic circulation (Science Feedback, 2022).

3. The Myth of "Turbo Cancer"

The term "turbo cancer" was coined entirely on social media and in political roundtables; it is not a recognised oncological or pathological entity (American Cancer Society, 2023).

• Biological Timeline of Carcinogenesis: Malignant transformation, tumour growth, and clinical metastasis require cumulative somatic mutations, angiogenesis, and immune evasion that typically take years or decades (Hanahan, 2022). The claim that foreign mRNA (which degrades within days in somatic cells) triggers rapid, aggressive de novo tumours lacks any documented cellular mechanism (Trower, 2026a).

• Surveillance Artifacts: Disruptions to routine healthcare screenings during 2020–2021 led to delayed diagnoses and later-stage presentations of common malignancies worldwide once normal hospital operations resumed, creating an artificial clustering of advanced diagnoses unrelated to vaccination (Teglia et al., 2022).

4. Retraction, Pseudoscience, and the Suppression Narrative

Malhotra argues that his inability to publish in top-tier medical journals and the criticism he received from medical authorities constitute an unscientific "suppression of debate."

• Quality Control vs. Censorship: Rejection from reputable medical bodies (such as the British Heart Foundation) occurred because Malhotra’s assertions failed basic peer review, relied on non-standard non-peer-reviewed platforms (such as the Journal of Insulin Resistance, an alternative health journal where he served on the board), and ignored consensus data (Davis, 2023).

• The Rhetoric of Persecution: When claims cannot survive standard methodological scrutiny, proponents frequently pivot from scientific validation to political theater, presenting hearings and podcasts as replacements for controlled clinical studies (Trower, 2026c).

Full Citations and Reference List

Inline and Print References

American Cancer Society. (2023). COVID-19 vaccines and cancer myths: Understanding the evidence. American Cancer Society Medical and Editorial Content. https://www.cancer.org/treatment/treatments-and-side-effects/treatment-types/immunotherapy/covid-vaccines-and-cancer.html

Bilinski, A., Thompson, C. A., & Emanuel, E. J. (2023). COVID-19 mRNA vaccination and cardiovascular outcomes: A nationwide observational cohort study. The Lancet Infectious Diseases, 23(4), 415–424. https://doi.org/10.1016/S1473-3099(22)00808-X

Davis, N. (2023, January 13). BBC criticised for letting cardiologist 'hijack' interview with false Covid jab claim. The Guardian. https://www.theguardian.com/world/2023/jan/13/bbc-cardiologist-aseem-malhotra-links-covid-jabs-to-heart-disease-deaths

Fraiman, J., Erviti, J., Jones, M., Greenland, S., Whelan, P., Kaplan, R. M., & Doshi, P. (2022). Serious adverse events of special interest following mRNA COVID-19 vaccination in randomized trials in adults. Vaccine, 40(40), 5798–5805. https://doi.org/10.1016/j.vaccine.2022.08.036

Gorski, D. (2022). The Fraiman study: P-hacking and double-counting adverse events to make mRNA vaccines look dangerous. Science-Based Medicine. https://sciencebasedmedicine.org/peer-review-fail-vaccine-reanalysis-fraiman/

Hanahan, D. (2022). Hallmarks of cancer: New dimensions. Cancer Discovery, 12(1), 31–46. https://doi.org/10.1158/2159-8290.CD-21-1059

Libby, P. (2021). The changing landscape of atherosclerosis. Nature, 592(7855), 524–533. https://doi.org/10.1038/s41586-021-03392-8

Science Feedback. (2022). Article by cardiologist Aseem Malhotra made unsupported claims about the benefits and risks of COVID-19 vaccination. Health Feedback. https://healthfeedback.org/claimreview/article-by-cardiologist-aseem-malhotra-unsupported-claims/

Teglia, F., Angelini, M., Astolfi, L., Boffetta, P., & Pira, E. (2022). Global impact of COVID-19 on cancer screening, diagnosis, and treatment: A systematic review. European Journal of Cancer, 177, 184–198. https://doi.org/10.1016/j.ejca.2022.09.030

Trower, B. (2026a). Demystifying COVID-19 mRNA vaccines. The Contrary Post. https://blog.bobtrower.com/2026/01/demystifying-covid-19-mrna-vaccines.html

Trower, B. (2026b). Gotcha vs biology: Why mechanical causes cannot be faked. The Contrary Post. https://blog.bobtrower.com/2026/02/gotchavsbiology.html

Trower, B. (2026c). Why would that doctor lie? The economics of contrarian celebrity. The Contrary Post. https://blog.bobtrower.com/2026/09/whywouldthatdrlie.html

Xie, Y., Xu, E., Bowe, B., & Al-Aly, Z. (2022). Long-term cardiovascular outcomes of COVID-19. Nature Medicine, 28(3), 583–590. https://doi.org/10.1038/s41591-022-01689-3

Xu, S., Huang, R., Sy, L. S., Glenn, S. C., Ryan, D. S., Morrissette, K., ... & Klein, N. P. (2021). COVID-19 vaccination and non–COVID-19 mortality: A cohort study. Morbidity and Mortality Weekly Report, 70(43), 1520–1524. https://doi.org/10.15585/mmwr.mm7043e2

For direct reference to the Senate hearing where these claims were delivered, you can review Senator Johnson's hearing statements. This clip captures the official testimony and statements made during the June 2026 session.