Maximum Truth

Maximum Truth is a data-driven substack exploring diverse topics like AI advancements, behavioral shifts in society, immigration policies, health science, and political predictions. It utilizes data analysis to delve into contemporary questions, aiming to provide insights beyond mainstream narratives by leveraging the author's background in math and economics.

Artificial Intelligence Societal Behavior Immigration Policy Health and Science Political Analysis Technology Bias

The hottest Substack posts of Maximum Truth

And their main takeaways

The Dawn of Woke AI

189 implied HN points • 24 Feb 24
🕹 Technology AI Bias Ethics Competition Social media Artificial Intelligence
  1. Google's Gemini Advanced AI displayed bias by predominantly erasing European features in its generated images.
  2. The head of Google's AI team, Jack Krawczyk, has displayed strong political views, influencing the direction of the AI's bias.
  3. Competition in the AI industry offers hope for less biased alternatives to heavily politicized AI models like Google's Gemini Advanced.

AIs ranked by IQ; AI passes 100 IQ for first time, with release of Claude-3

126 implied HN points • 05 Mar 24
🕹 Technology AI Artificial Intelligence Chatbots
  1. AIs can improve their IQ scores when given special accommodations in IQ tests, similar to how blind individuals may require accommodations for certain tasks.
  2. Claude-3 represents a significant leap in AI intelligence, showing a consistent increase in IQ scores across different versions, prompting considerations of future AI advancements.
  3. AI rankings based on IQ reveal variations in intelligence among different AIs, with Claude leading the pack, followed by ChatGPT. The ranking can guide decisions on experimenting with different AIs.

Top AIs still fail IQ tests

118 implied HN points • 27 Feb 24
🔬 Science AI Artificial Intelligence Cognitive abilities Intelligence
  1. Top AIs still struggle with IQ tests, showing limitations in understanding logic and spatial patterns.
  2. AI's strength lies in database knowledge and pattern matching rather than general intelligence.
  3. Current AIs, like ChatGPT-4, perform only slightly better than random guessers on IQ tests, indicating a lack of generalized intelligence.

Border Crossing Data Show Incentives Matter

88 implied HN points • 12 Feb 24
🇺🇸 U.S. Politics Immigration Government Policy Media Debate
  1. Illegal migration is primarily determined by ease of entry, as shown by border crossing data
  2. Data indicates that incentives matter in influencing the number of migrant crossings, with policies like 'Remain in Mexico' impacting the flow
  3. The media tends to avoid in-depth discussions on the pros and cons of open borders, preferring a simpler 'don't be racist' narrative

Epicurus, the First "Rationalist"?

37 implied HN points • 15 Mar 24
📖 Philosophy Ancient Philosophy Moral Philosophy Epistemology Metaphysics Ethics
  1. Epicurus, an ancient philosopher, focused on maximizing pleasure and minimizing pain, inspiring utilitarianism.
  2. Epicurus had a rational and secular approach to understanding the world, advocating for scientific explanations over myths.
  3. He had insightful scientific views for his time, such as proposing the existence of atoms and explaining the origins of scents.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

High School Then and Now: Data show big shifts in behavior

101 implied HN points • 22 Jan 24
🚌 Education
  1. High school students are reading much less long-form content.
  2. High school students are dating much less, with about 50% of 12th graders never having dated.
  3. High school students are engaging in less adult and independent behavior like driving, going out, and getting enough sleep.

The AI Revolution Has Begun

71 implied HN points • 08 Dec 22
  1. AI is advancing rapidly and can now write eloquently and code at a basic level.
  2. AI can excel at strategy, negotiation, and even ethical guidance.
  3. Current AI struggles with critical thinking, relying on word association for conclusions.

DEEP DIVE: Election Betting vs 538

33 implied HN points • 28 Nov 22
  1. People have been interested in predicting the future throughout history.
  2. Election bettors are very accurate in predicting election outcomes within the 20% to 80% range.
  3. Betting odds slightly outperformed 538 in accuracy for the period 2016 - 2020, but 538 now holds a narrow lead after 2022.

DEEP DIVE: Should Naturally-Immune People Get Vaccinated?

50 implied HN points • 10 Sep 21
  1. Vaccinated people with previous COVID-19 have increased immunity. Getting vaccinated on top of natural immunity further reduces the chances of COVID-19.
  2. Natural immunity provides solid protection, especially in the long run, comparable to vaccines. However, vaccines offer an added layer of protection against COVID-19, making it beneficial for naturally-immune individuals to consider vaccination.
  3. Side effects of Pfizer and Moderna vaccines are not significantly worse for individuals who already had COVID-19. Both vaccines offer similar levels of protection in people with prior infection.

DEEP DIVE: Does Covid Cause Significant Brain Harm?

29 implied HN points • 15 Oct 21
  1. Brain scans show minor effects compared to normal brain changes over 3 years.
  2. Age plays a role, with older individuals experiencing a slightly more negative impact.
  3. Long Covid, especially affecting smell, is real but rare; media headlines may overstate its prevalence.

Humbly grappling towards truth

16 implied HN points • 07 Feb 21
  1. The blog focuses on digging into important questions often overlooked by mainstream media.
  2. The author's background in math and economics aids in investigating various subjects like tech, science, philosophy, and politics.
  3. The blog aims to provide non-ideological, data-driven content to help readers learn and move away from the noise of daily news.