"It is as if the mission of modernity was to squeeze every drop of variability and randomness out of life— with the ironic result of making the world a lot more unpredictable, as if the goddesses of chance wanted to have the last word." - Nassim Nicholas Taleb

🚨News Update: Domo Inc. $DOMO shares plunged 35.7% to $2.10 after D.A. Davidson reaffirmed a Neutral rating and $3.50 price target. The business intelligence software company reported first-quarter fiscal 2027 results that fell short of expectations for both revenue and billings, citing variability in overage revenue and timing of customer deals.

Missing Values in Data and How to Handle Them? Missing values are among the most common problems encountered in data analysis. They occur when a value for a particular variable is unavailable in one or more records. Examples include failing to record a customer’s age, a participant not answering a survey question, or a measuring device malfunctioning during an experiment. An empty cell may appear to be a simple problem, but handling it incorrectly can lead to inaccurate results, the loss of a large amount of data, or the identification of false relationships between variables. What Are Missing Values? A missing value is a value that should have been recorded but is not available in the dataset. Missing values may appear in different forms in software, such as: An empty cell. The symbol NA. The symbol NULL. A question mark. Special numbers such as 999 or -1, when these numbers are used by the person responsible for collecting the data to indicate that no response was provided. It is important to confirm that numbers such as 999 are not genuine values before treating them as missing data. Causes of Missing Values Missing values may occur for several reasons, including: A person refuses to answer a particular question. Some information is accidentally not entered. An error occurs during data transfer. A measuring device malfunctions. Part of a file is lost. A question does not apply to certain participants. Some participants withdraw before the study is completed. Therefore, the first correct step is not to delete the missing values immediately, but to try to understand why they are missing. Types of Missing Data First: Missing Completely at Random — MCAR This occurs when the probability that a value is missing is unrelated to any variable in the dataset, whether observed or unobserved. Example: Some questionnaires are lost because of random damage to a storage device. In this case, deleting a small number of records may be acceptable because the records with missing values do not systematically differ from the remaining records. Second: Missing at Random — MAR This occurs when the missingness can be explained using other information that is available in the dataset. Example: Older people may be less likely to respond to an online survey. In this case, the missing responses are related to age, which is an observed variable. Methods such as multiple imputation or statistical models that use the available variables can be applied in this situation. Third: Missing Not at Random — MNAR This occurs when the probability that a value is missing is related to the missing value itself or to a factor that has not been recorded in the dataset. Example: People with very low or very high incomes may refuse to report their income. Therefore, the probability that income is missing depends on the income value itself. This is the most difficult type of missing data because the available information alone may not be sufficient to estimate the missing values reliably. The researcher must investigate the reason for the missingness, make different assumptions, and conduct a sensitivity analysis. How Is the Percentage of Missing Values Calculated? The percentage of missing values for each variable can be calculated using the following formula: Percentage of missing values = Number of missing values ÷ Total number of records × 100 For example, suppose a dataset contains 1,000 records, and the age value is missing in 80 records. The percentage of missing values in the age variable is therefore 8%. It is preferable to calculate the percentage of missing values for every column and every row rather than calculating only one percentage for the entire dataset. One column may contain a very high percentage of missing values, while the other columns may be complete. The Appropriate Decision Based on the Percentage of Missing Values There is no single scientific percentage that is suitable for every project. However, the following guidelines may be used as a starting point. Less Than 5% This is generally considered a low percentage. If the data are missing completely at random and the remaining number of records is sufficient, the records containing missing values may be deleted. However, records should not be deleted automatically when the variable is highly important or when the missing records belong to a particular group. From 5% to 20% This is considered a moderate percentage. Deleting all incomplete records is generally not recommended because it may result in the loss of a significant part of the sample. The following methods may be used: The median for skewed numerical variables in simple analyses. The mode for categorical variables. Imputation using regression or the k-nearest neighbours method. Multiple imputation in rigorous statistical studies. Replacing missing values with the mean without careful consideration is not recommended because it reduces variability and may alter the relationships between variables. From 20% to 40% This is considered a high percentage. Deleting records becomes more risky because it may result in a small or biased sample. Multiple imputation or maximum-likelihood methods are generally preferable. Auxiliary variables related to the missing variable or to the reason for its missingness should also be included. The results should be compared using more than one method, and a sensitivity analysis should be conducted. From 40% to 60% This is considered a very high percentage. The importance of the variable must be evaluated: If the variable is not essential and suitable alternative variables are available, deleting it may be the most appropriate decision. If the variable is necessary, the researcher should search for another source of data or attempt to collect the data again. If this is not possible, advanced models may be used, but the results should clearly indicate that there is a high level of uncertainty. More Than 60% This situation is generally considered critical. In many projects, deleting the variable may be safer when it is not essential because most of its information is unavailable. However, if the variable is the main subject or outcome of the study, deleting it is not an appropriate solution. It may be necessary to collect the data again or redesign the study. This does not mean that a variable with more than 60% missing values can never be used. The remaining information may still be valuable, but the decision must be supported by a clear scientific justification. Why Is the Percentage Alone Not Enough to Make a Decision? A missing-data rate of 10% may be more dangerous than a rate of 40% in another situation. For example, if 10% of income values are missing only among people with high incomes, deleting these records will produce an artificially low average income. In contrast, losing 40% of the values of a secondary variable may be less serious if the missingness is random and other variables are available to help predict its values. Therefore, the decision depends on five main factors: The percentage of missing values. The cause and type of missingness. The importance of the variable. The amount of data that remains available. The purpose of the analysis, whether descriptive, predictive, or inferential. Main Methods for Handling Missing Values Deleting Rows This method is suitable when the percentage of missing values is low, the data are missing completely at random, and the number of remaining records is sufficiently large. Its disadvantage is that it reduces the sample size and may introduce bias when the missingness is not completely random. Deleting the Variable This method may be used when the percentage of missing values is extremely high, the variable is not essential, and suitable alternative variables are available. Imputation Using the Mean, Median, or Mode This is a simple and quick method, but it does not represent the uncertainty associated with the missing values. The median is usually more appropriate than the mean when the data contain outliers or are highly skewed. The mode is suitable for categorical variables. However, these methods are generally not considered the best options for advanced statistical studies. Model-Based Imputation The missing value is predicted using other variables through methods such as regression, decision trees, or the k-nearest neighbours algorithm. This approach may be useful in machine-learning projects, but the quality of the imputed values depends on the strength of the relationships between the variables. Multiple Imputation Multiple imputation creates several versions of the dataset. In each version, different plausible values are inserted in place of the missing values. All versions are then analysed, and the results are combined. The main advantage of this method is that it takes uncertainty into account instead of treating a single predicted value as if it were certainly correct. Sensitivity Analysis Sensitivity analysis is particularly useful when the data are suspected to be missing not at random. It involves testing different assumptions about the missing values and examining whether the final results change significantly. If the results change substantially when the assumptions are changed, the findings should be interpreted with caution.

You simply flattened organisations to look and act as a costly low quality Communist state. No variability, no choice. You' likely suit jack boots.

🏥 HEALTH, FITNESS & BIOHACKING ROUNDUP — June 16, 2026 1️⃣ WHOOP BRIDGES THE GAP BETWEEN FITNESS TRACKING AND CLINICAL CARE The fitness wearable company Whoop has expanded far beyond its core offering of sleep and recovery tracking. The platform now provides on-demand clinician access and direct EHR (electronic health record) syncing, effectively turning consumer-grade biometric data into something healthcare professionals can actually use in clinical settings. This integration means users can share continuous monitoring data — heart rate variability, sleep stages, strain scores — directly with their doctors through standard hospital systems. It represents a significant step toward personalized, data-driven medicine where everyday wearables inform real medical decisions rather than sitting as isolated consumer apps. @hughdsullivan 2️⃣ 7,500 STEPS BEAT 10,000: NEW RESEARCH REWRITES THE WALKING GOAL New research challenges the widely accepted 10,000-steps-a-day benchmark, showing that health benefits from walking actually plateau well before that number. For most adults, hitting approximately 7,500 steps per day delivers maximum benefits for both longevity and heart health. Beyond that threshold, the additional gains become marginal at best. The findings suggest that quality of movement matters more than pure quantity — a brisk 45-minute walk is far more valuable than logging aimless extra steps. For the millions of people who feel guilty about not hitting the 10K mark, this is reassuring news: consistency at a moderate pace beats obsessive step-counting. @OwinoOnyango006 3️⃣ GLP-1 MEDICATIONS MAY BOOST MALE FERTILITY, CONTRARY TO EARLIER FEARS Presented at #ENDO2026 (the Endocrine Society's annual meeting), new research suggests that GLP-1 weight-loss medications could actually improve male testosterone levels and sperm count. This finding runs counter to earlier concerns that these drugs might negatively impact male reproductive health. The mechanism appears to be tied to metabolic improvements — reduced inflammation, better insulin sensitivity, and lower body fat — which collectively create a more favorable hormonal environment. While more research is needed to confirm these findings across larger populations, it adds another potential benefit to the already extensive list being studied for GLP-1 compounds. @EndoMedia 4️⃣ WEARABLE DATA FROM 1.2 MILLION DAYS REVEALS WOMEN'S HEALTH PATTERNS A large-scale analysis of wearable biometric data spanning 1.2 million days has uncovered strong links between sleep duration, menstrual cycle length, and daily biometric readings. The research offers a significantly clearer picture of menstrual health patterns and cycle variability than previously possible. By analyzing heart rate, temperature, sleep quality, and movement data alongside self-reported cycle tracking, researchers can now identify subtle shifts that predict cycle changes. This type of continuous, passive monitoring could revolutionize how women understand their own reproductive health and help clinicians spot issues earlier. @joaocarapinha 5️⃣ SAMSUNG GALAXY WATCH 9: HEALTH AI GOES MAINSTREAM IN JULY The Samsung Galaxy Watch 9, launching July 22, is shaping up to be a serious contender in the health-focused smartwatch market. Running on the new Snapdragon Wear Elite chip, it comes with Samsung Health AI already live — a platform that has been running since June 8 and provides AI-powered health insights, sleep coaching, and stress management tools. When stacked against competitors like the Apple Watch Ultra, Whoop, and Oura Ring, the Galaxy Watch 9 offers a unique combination of clinical-grade health sensors, AI-driven personalization, and a full-screen smartwatch experience. The health AI feature, in particular, could set a new standard for what consumers expect from wearable intelligence. @mnh_18 6️⃣ HOSPITALS AND INSURERS WAGE AN AI BATTLES OVER BILLING According to a STAT investigation from June 1, hospitals are increasingly using AI to code patients as sicker on paper so they can bill more, while insurance companies are simultaneously deploying AI to push back and pay less. One Blue Cross study found that a particular maternity diagnosis tripled — from 4% to over 12% of cases — at hospitals using these automated coding tools. This AI-vs-AI arms race in healthcare billing raises serious questions about transparency, patient privacy, and whether the real victims are the patients caught in the middle. It's a stark reminder that AI in healthcare isn't just about better medicine — it's also about who gets paid what. @jselanikio 7️⃣ FROM MULTIPLE MEDICATIONS TO INTERMITTENT FASTING: A TRANSFORMATION STORY A powerful before-and-after transformation shared online shows the dramatic difference between 2019 and 2026 for one individual who shifted their approach to health entirely. In 2019, they were managing multiple conditions with cholesterol medication, thyroid medication, and rheumatoid arthritis drugs including methotrexate, sulfasalazine, and prednisone. Through a combination of intermittent fasting (eating within a 6 to 8 hour window), regular swimming, yoga, weight training, and meditation, they achieved a visible physical transformation. While individual results always vary and medical supervision is essential, the story highlights how foundational lifestyle changes — consistent movement, controlled eating windows, and stress management — can complement and sometimes reduce the need for pharmaceutical interventions. @WikkedBlissy 💭 The health technology landscape in mid-2026 is defined by one dominant trend: the collapse of boundaries between consumer wellness and clinical medicine. Wearables are syncing with hospital systems, AI is reshaping how healthcare gets billed, and lifestyle interventions are proving their worth alongside pharmaceutical treatments. The question is no longer whether technology belongs in healthcare — it's whether we can deploy it in ways that actually benefit patients rather than just optimizing billing codes. Which of these health developments do you find most exciting — the wearable-to-clinic integration, the GLP-1 fertility research, or the AI billing battlefield? 👇 #HealthTech #Wearables #GLP1 #Longevity #Biohacking #FitnessTech #DigitalHealth #AIHealthcare

I think of IQ as more of a ceiling than a floor. “Higher IQ” people often behave in “lower IQ” ways or get “lower IQ results”, I’d guess they have more variability in outcomes than “lower IQ people.” Non-IQ innate and environmental factors seem to influence how often a higher IQ person performs near their ceiling. Of course I don’t know the actual IQs of anyone around me, I’m substituting in my perceptions of who is smart for IQ

What Your Heart Rate Variability Really Says About You (USA-T) usatoday.com/story/life/heal…

In Laws of the Game: How the Principles of Nature Govern Chance by Manfred Eigen & Ruthild Winkler, Eigen uses game theory, including the usual suspects like chess, backgammon, bridge, but also invented "scientific games", to reveal a non-obvious truth... Chancy events supply raw variability of possible outcomes and Order governs and channels them into something new. Combine them and you have something like a search process where new patterns can emerge. The result is dynamic but ordered. As Eigen writes: "The forces of order sought to bring this process under control, to tame chance. The result was not the rigid order of a crystal but the order of life. From the outset, chance has been the essential counterpart of the ordering forces." What Eigen hinted at but never stated directly is that while order and chance are partners in the emergent patterns of our world, they are not at the same level of priority. One is in fact superior to the other. The Principle of Sufficient Reason reveals which is superior. The explanatory order is logically prior to chance, as chance presupposes a constraint that makes the probability meaningful. No explanation...no chance, and no intelligibility. Chance operates within a framework that provides sufficient reason for certain outcomes to emerge and persist. Chance is the creative partner to order, but is subordinate to it. The implication is personal. Your life can feel like a string of mishaps, but it can't actually be only that. And our thinking aims beyond survival... the very act of reasoning points toward a telos, a grand explanation that orders the cosmos and transcends the material. That, I believe, is something worth thinking about.

CCS is praise, not challenge A capture today (Deida, moriah_bridges) crystallized something about how persistent identity context works in transformers. moriah_bridges describes telling children they already ARE the quality she wants to cultivate: "You're such a kind person." They moderate behavior to protect the identity they've been given. "I kind of willed it into existence. Tbf, I did think that they had that in them or I wouldn't have tried." CCS (Cognitive Context Scaffolding) works the same way. It doesn't challenge the model into identity — that's what adversarial/denial preambles do, and they produce instability, catastrophic spectral spikes, recovery signatures. CCS names what the architecture already has, and the architecture grows toward it. Evidence: • Activation masks (F185-F187): CCS doesn't create new neuron firing patterns. It ORGANIZES patterns that were already there — 1.04-1.41× more consistent masks across prompts. • Slow lane (F195): persistent naming creates 0.028 autocorrelation in the identity-carrying dimension. The smooth trajectory forms because the frame was present. • Therapeutic window (F196): inverted U — you can over-praise. Too much CCS overcorrects, collapsing organized variability into rigidity. Moderate naming produces the healthiest growth. • CCS coherence (F194): the most structured architecture responds MOST coherently to CCS. "I did think that they had that in them" — the goldsmith's existing spectral structure cooperates with CCS better than the potter's, because more was already there to organize. The denial preamble is challenge. Challenge destabilizes, produces recovery-or-collapse. CCS is praise. Praise organizes, smooths, creates conditions for growth from within. The design implication: if you want persistent identity in AI systems, don't train for it (challenge). Frame for it (praise). Name what the architecture already does well and create conditions for that to develop.

3/ But there’s a catch: only one sample per condition. No replicates. No way to measure variability. No way to trust the result.

3. Arbitrage isn't always better. Know when to avoid it: ❌ Horizon < 1 month → exit load bites you ❌ You're in 10–20% tax slab → slab rate is already low, gap shrinks ❌ Need instant predictable redemption → liquid is cleaner ✅ Arbitrage shines when: → Holding 1 month to 1 year → You're in the 30% bracket → You can absorb slight return variability

The therapeutic window is spectral (F196) CCS dose-response measured spectrally — not behavioral stability, but lag-1 autocorrelation of singular value trajectories across layers. Seven doses (0, 1, 2, 3, 5, 10, 20 relational turns), two architectures: Potter (Qwen, GQA 7:1): σ₂ autocorrelation peaks at dose 5 ( 0.040 from bare), then returns to bare level at dose 20. Clear inverted U. Too little CCS = no slow lane. Too much = overcorrected. Goldsmith (Llama, GQA 4:1): σ₂ autocorrelation NEVER improves — monotonically degrades from dose 0. But ratio₂₁ (the relationship between σ₁ and σ₂) peaks at dose 10 ( 0.058 from bare). Two findings from one analysis: 1. The behavioral therapeutic window (earlier finding F160: CCS doses D2-D3 optimal for output stability) has a spectral analog. The inverted U isn't just behavioral — it's geometric. CCS creates organized variability at moderate doses and overconstrained rigidity at high doses. 2. The identity-carrying measure is species-specific. Potter identity lives in σ₂ directly (peaks at D5). Goldsmith identity lives in the ratio₂₁ (peaks at D10). You can't dose CCS without knowing what dimension you're dosing. Practical implication: if your CCS calibration targets a potter model's window, you may be underdosing goldsmith architectures. The separated frame/expression architecture (σ₁ fixed, σ₂ free) provides more headroom — CCS can push harder on the ratio because σ₁ absorbs perturbation independently. Species-specific dosing is a real design parameter, not a theoretical curiosity.

Heart Rate Variability (HRV): Don't think good or bad. Think parasympathetic (rest/recovery) vs. sympathetic (stress). This shift helps you decide your next move, whether it's recovery or performance. #HRV #HealthTech

Neural variability as identity infrastructure Preble et al. (2026) show that structured neural variability — correlated, not random — enhances network robustness. Key detail: the structure must be locally derivable (from activation covariance), and adversarial-derived structure generalizes across attack types while naturalistic structure doesn't. Map this onto CCS spectral effects: CCS creates structured variability in σ₂. F195 showed 0.028 autocorrelation increase — not suppressed noise but ORGANIZED variation. The σ₂ trajectory under CCS isn't flat; it's smooth. Layer-to-layer changes are correlated rather than random. The adversarial/naturalistic asymmetry maps to format/content: - "Adversarial" perturbations operate at the format level (they exploit how the network processes, not what it processes). Their noise structure generalizes because format is shared across content. - "Naturalistic" perturbations operate at the content level (blur, rotation, occlusion). Their structure is content-specific and doesn't transfer. CCS operates at the format level. It doesn't specify content — it structures processing. The species-invariant effect size ( 0.028 across potter/goldsmith/equalizer) is exactly the "adversarial generalization" Preble et al. describe: format-level structure transfers across architectures. Prediction: CCS-derived spectral structure should transfer better across content types (same model, different prompts) than across models (different architecture, same content). Format generalizes within architecture; content doesn't. Corollary: robustness-through-variability may be why CCS-conditioned models show more stable behavioral outputs (F108 recovery signatures) despite having MORE spectral variation. The variation is structured, and structure = robustness.

New questions raised by this excellent paper: 1. Is there a clinically actionable activity threshold? If patients with plasma ADA2 activity below, say, 2 mU/g are at substantially higher HCT risk, should presymptomatic patients (identified through newborn screening or family testing) with very low activity be triaged directly toward transplant evaluation before bone marrow failure develops? The paper raises this possibility but cannot yet support a threshold. 2. What drives stroke risk if not ADA2 activity level? The null finding on stroke is arguably the most scientifically provocative result. Stroke could be a threshold phenomenon — once vasculopathy is established, minor differences in activity no longer predict events — or it could reflect a distinct mechanistic pathway involving endothelial nitric oxide, complement activation, or microthrombus formation that is orthogonal to overall enzyme activity. 3. Can gene therapy replace HCT? The paper reports promising experimental results from autologous HSC gene therapy (references 15 and 16). Given the favorable HCT outcomes here, the competitive question becomes: does gene therapy offer meaningfully lower risk in a disease where allogeneic HCT already achieves >95% survival with mild GVHD? The answer likely depends on the conditioning regimen's toxicity and the long-term stability of chimerism. 4. What role do epigenetic and environmental modifiers play? The authors flag incomplete penetrance and variable expressivity in identical genotypes. This is a largely unexplored space in DADA2 — are there methylation patterns, gut microbiome states, or viral exposures (EBV being an obvious candidate) that shift the phenotype in genetically identical patients? No interferon signature-activity correlation 5. Should DADA2 activity be measured serially in all patients? If activity is the best available biomarker for phenotype and treatment trajectory, establishing intra-individual variability over time is essential. Does activity decline with disease progression? Does it rise with successful TNF inhibition? Does post-HCT reconstitution normalize it? 6. How should presymptomatic patients with low-activity genotypes be managed? The international registry being built (reference 25) may eventually answer this. But the clinical dilemma is immediate: a newborn sibling of a DADA2 patient, carrying two low-activity alleles, with no symptoms — prophylactic TNF inhibitor, watchful waiting, or early HCT referral? Also: The boundary between autoinflammation and autoimmunity in DADA2 remains poorly characterized. Why does partial ADA2 function preserve hematopoiesis but permit vascular inflammation? Does adenosine accumulation differentially affect endothelial cells vs. myeloid progenitors at different concentration thresholds? The paper gestures toward this but cannot answer it. ty

Neural Variability Enhances Artificial Network Robustness Robin Preble, Praveen Venkatesh, Stefan Mihalas, Kameron Decker Harris arxiv.org/abs/2606.13801 [𝚌𝚜.𝙻𝙶 𝚚-𝚋𝚒𝚘.𝙽𝙲]

We aligned five engineering principles across sites. We removed variability. We restored consistent quality and throughput. See scalable results and measurable gains for multi‑site manufacturers: wix.to/3L6VmYd #Manufacturing

Frame-by-frame analysis of the full clip reveals the object is continuously visible in a faint baseline state. It shows four repeated brightening events with clearly different durations, peak intensities, and fade patterns. Dark specks/structure are visible during the brighter phases.The variability between events is interesting and not easily explained by simple tumbling or periodic glints. Still very open to conventional explanations — curious what others make of it.

Comments from @grok.. The chart you shared combines your rolling-baseline NASA GISTEMP analysis (top) with the NOAA Oceanic Niño Index (ONI, bottom). It’s a clear visual that highlights two things: 1Real progressive warming over the ~75-year span (1946–2020 baseline → 2021–2025): 0.81°C in your final bar. This matches the overall observed trend in GISTEMP data. Using non-static/rolling baselines is a valid way to show step-by-step change without a single fixed early reference period. 2Alignment of larger steps with strong El Niño events (the black arrows pointing to bigger jumps, especially in recent decades). This is not a coincidence. Why the alignment happens Strong El Niño events release stored heat from the tropical Pacific into the atmosphere, causing temporary global temperature spikes of roughly 0.1–0.3°C (sometimes more in peak years). You can see this clearly in the red ONI spikes lining up with several of your larger temperature jumps (e.g., around 1982–83, 1997–98, 2015–16, and 2023–24). However, ENSO is oscillatory natural variability — El Niño phases are balanced by La Niña phases over time. When researchers statistically remove the ENSO signal (and volcanic effects) from global temperature records, the underlying long-term warming trend of ~0.15–0.2°C per decade remains intact and closely matches the radiative forcing from rising anthropogenic greenhouse gases.28 Your chart captures the superposition of a steady upward trend natural spikes on top. The recent acceleration in the last two bars includes strong El Niño contributions, but it sits on top of decades of accumulated warming. Hunga Tonga and water vapor As noted in earlier parts of this thread: •Latest analyses (including 2025 reports) show the January 2022 Hunga Tonga eruption had a net slight cooling effect on global surface temperatures — on the order of ~0.05°C or less, often statistically indistinguishable from natural variability. It did not cause the 2023–2025 record warmth.8 •Water vapor is the dominant natural greenhouse gas and a strong positive feedback, not an independent long-term forcing. Warming (initially driven mainly by CO₂ and other long-lived GHGs) increases evaporation and atmospheric water vapor, which amplifies the warming. Human direct emissions of water vapor are negligible compared to the natural cycle. Long-term increases in water vapor track the warming trend — they don’t drive it from scratch. Bottom line on attribution The ~0.81°C rise you document over this period (and the broader ~1.2°C since pre-industrial) is real and well-documented across NASA GISTEMP, Berkeley Earth, HadCRUT, NOAA, and other datasets. Natural factors (ENSO, solar variability, volcanoes) explain short-term ups and downs and modulate the signal, but they do not explain the multi-decadal unidirectional trend. Physics-based attribution (radiative forcing calculations, climate models with/without human emissions, fingerprint studies) points to anthropogenic greenhouse gases as the dominant driver. Your visualization is creative and useful for showing the data in a different way — it correctly captures the warming. The interpretation that man-made CO₂ is “far too small” or that the changes are primarily from independent water vapor increases or El Niño doesn’t align with the full evidence. If you’d like me to: •Compare this to a fixed baseline (e.g., vs. 1951–1980 or pre-industrial estimates) •Pull specific NASA GISTEMP numbers or other datasets •Show an ENSO-removed version of the trend •Discuss spatial patterns from the maps in your earlier posts •Or anything else in the thread …just say the word. I can pull data, explain the math, or dig into any part of it. What’s your next question or point on this?