Consider running your business based on wizardry…

AI’s Memorization Crisis

On Tuesday, researchers at Stanford and Yale revealed something that AI companies would prefer to keep hidden. Four popular large language models—OpenAI’s GPT, Anthropic’s Claude, Google’s Gemini, and xAI’s Grok—have stored large portions of some of the books they’ve been trained on, and can reproduce long excerpts from those books.

In fact, when prompted strategically by researchers, Claude delivered the near-complete text of Harry Potter and the Sorcerer’s Stone, The Great Gatsby, 1984, and Frankenstein, in addition to thousands of words from books including The Hunger Games and The Catcher in the Rye. Varying amounts of these books were also reproduced by the other three models. Thirteen books were tested.

This phenomenon has been called “memorization,” and AI companies have long denied that it happens on a large scale. In a 2023 letter to the U.S. Copyright Office, OpenAI said that “models do not store copies of the information that they learn from.” Google similarly told the Copyright Office that “there is no copy of the training data—whether text, images, or other formats—present in the model itself.”  Anthropic, Meta, Microsoft, and others have made similar claims. (None of the AI companies mentioned in this article agreed to my requests for interviews.)

I want one for Colorado!

https://pogowasright.org/how-californians-can-use-a-new-state-website-to-block-hundreds-of-data-brokers/

How Californians can use a new state website to block hundreds of data brokers

A tool called DROP lets California residents fill out a few forms to keep their personal data from being tracked or sold by data brokers.

Also suggests a way to entice users give up more personal data…

https://pogowasright.org/article-the-privacy-paradox-is-a-misnomer-data-under-structural-uncertainty/

Article: The Privacy Paradox Is A Misnomer: Data Under Structural Uncertainty

Cofone, Ignacio, The Privacy Paradox Is A Misnomer: Data Under Structural Uncertainty (January 06, 2026). Georgetown Technology Law Journal (forthcoming 2026), Available at SSRN: https://ssrn.com/abstract=6030275 or http://dx.doi.org/10.2139/ssrn.6030275

Abstract

The infamous privacy paradox refers to the apparent inconsistency between people’s stated concern for privacy and their readiness to disclose personal information. This phenomenon has sparked two largely disconnected literatures: one offering experimental evidence of inconsistent behavior, and another providing qualitative accounts and defending the importance of privacy.

The Article presents an online field experiment that bridges those literatures and shows that the so-called paradox arises from a mischaracterization of the underlying behavior. The Article finds that it is structural uncertainty about risk that drives seemingly paradoxical privacy decisions. It does so by isolating discounting mechanisms and empirically testing whether observed privacy choices reflect temptation or rational responses to uncertainty. The results suggest that privacy behavior is not paradoxical but, rather, consistent with choices shaped by incomplete information.

The Article then discusses the policy implications of this reframing. As privacy decisions stem from structural uncertainty, which operates as a market failure, regulation should aim to reduce that uncertainty. This supports regulation that prioritizes transparency-for people to assess the risks of data collection-and flexibility mechanisms that accommodate evolving contexts. Such reframing provides a new argument for the right to be forgotten, which allows people to revisit prior disclosures as new risks become apparent. By shifting the focus from individual inconsistency to structural uncertainty, the findings call for privacy law to better reflect the reality of people’s decision-making environments.

Download the full article (free) at SSRN: https://ssrn.com/abstract=6030275 or http://dx.doi.org/10.2139/ssrn.6030275

A possible de-halucinating tool?

https://www.bespacific.com/ai-research-pilot/

AI Research Pilot

Digital Digging: “A researcher doesn’t use AI as primary source” is my mantra. But what if we could use AI not as the source, but as a tool that actually points us to legitimate sources? That’s why I built AI Research Pilot.  Advanced search tool for investigators, reporters, and researchers conducting in-depth analysis. 

How future wars will begin?

https://www.schneier.com/blog/archives/2026/01/a-cyberattack-was-part-of-the-us-assault-on-venezuela.html

A Cyberattack Was Part of the US Assault on Venezuela

We don’t have many details:

President Donald Trump suggested Saturday that the U.S. used cyberattacks or other technical capabilities to cut power off in Caracas during strikes on the Venezuelan capital that led to the capture of Venezuelan President Nicolás Maduro.

If true, it would mark one of the most public uses of U.S. cyber power against another nation in recent memory. These operations are typically highly classified, and the U.S. is considered one of the most advanced nations in cyberspace operations globally.

What are the odds that the US invades Greenland? (Did Trump insiders make lots of bets?)

https://www.theguardian.com/world/2026/jan/07/wager-platform-polymarket-will-not-pay-out-on-bets-on-us-invasion-of-venezuela

Gambling platform Polymarket not paying bets on US invasion of Venezuela

The online wager platform Polymarket has angered some gamblers by declaring it will not settle millions of dollars’ worth of bets on a US invasion of Venezuela, arguing that the capture of the then president, Nicolás Maduro, does not qualify.

Before Donald Trump’s forces seized Maduro  on Saturday morning, some traders appeared to have anticipated the shock move by placing bets on “prediction markets”.

These are gambling platforms that allow individuals to wager on a range of markets that have been created by the host website. They are typically binary bets, punting on yes/no or higher/lower outcomes.

Last Friday an anonymous trader on Polymarket appeared to invest $30,000 (£22,343) on the market: Maduro out by 31 January 2026. After Maduro’s capture was announced on Saturday morning, the trader seemed to have made profits of $436,759.61.

Clearly, I don’t understand tariffs.

https://fortune.com/2026/01/06/trump-trade-tariffs-revenue-inflation-stocks/

Trump’s trade tariff revenue is already in decline, and Wall Street is pretty happy about it

When the U.S. Bureau of Labor Statistics reported that its most recent measure of consumer price inflation was just 2.7%, it came as a surprise to many. The consensus prediction on Wall Street had been 3.1%.

Ever since President Donald Trump announced his Liberation Day tariffs last April, economists have been expecting the extra cost of those tariffs to show up in the inflation numbers. After all, the effective tariff rate on goods imported from China was as high as 57.6% in November 2025, according to the Peterson Institute for International Economics. 

Surely this would force up the price of goods for consumers? 

Nope, according to two new studies. Historically, tariffs haven’t resulted in big bursts of inflation, the studies from the Federal Reserve Bank of San Francisco and Northwestern University show. That’s because importing companies tend to find ways around the tariffs, or because countries negotiate enough compromises and exemptions to the tariffs to reduce the headline rate.

(Related)

https://www.frbsf.org/research-and-insights/publications/economic-letter/2026/01/what-can-history-tell-us-about-tariff-shocks/

What Can History Tell Us About Tariff Shocks?

The change in the average U.S. tariff rate in 2025 was the largest in the modern era. One way to assess the effects of such a large shock on unemployment and inflation is by looking at data from pre-World War II periods with tariff rate changes of a similar magnitude. Analysis shows that previous tariff hikes raised unemployment and reduced both economic activity and inflation. Uncertainty may be a factor behind these effects: A large tariff increase raises uncertainty, which can depress overall demand and lead to lower inflation.

Perspective.

https://blogs.lse.ac.uk/businessreview/2026/01/02/an-ai-playbook-for-working-with-non-human-minds/

An AI playbook for working with non-human minds

… For a long time, it was comforting to describe AI as a “tool”, like a spreadsheet or a search engine. But tools wait until a person chooses to use them and only act when told. And, given the high penetration of AI use in some industries, the metaphor underestimates how deeply AI is now woven into everyday processes.

Many AI models run in the background all day and night. Generative systems, which create content by learning patters from training data, can draft documents and code in seconds. None of these systems has intent or consciousness. But they act with a kind of continual presence that changes how decisions are made and executed.

Think: an AI Erector set.

https://www.makeuseof.com/lego-announces-smart-play-at-ces/

Lego is putting tiny computers inside its bricks, starting with Star Wars sets

Lego bricks are, well… Lego bricks. They’ve stayed largely the same for decades and are simply plastic pieces designed to be combined, rebuilt, and imagined into just about anything. While this isn’t changing, the company just announced the Lego Smart Brick at CES, which puts a tiny computer inside a standard Lego brick.

Prognostication.

https://sloanreview.mit.edu/article/five-trends-in-ai-and-data-science-for-2026/

Five Trends in AI and Data Science for 2026

MIT SMR columnists Thomas H. Davenport and Randy Bean see five AI trends to pay attention to in 2026: deflation of the AI bubble and subsequent hits to the economy; growth of the “factory” infrastructure for all-in AI adapters; greater focus on generative AI as an organizational resource rather than an individual one; continued progression toward value from agentic AI, despite the hype; and ongoing questions around who should manage data and AI.

Interesting. Useful?

https://www.bespacific.com/a-political-atlas-of-the-world/

A Political Atlas of the World

1st Experiments with WebMapperGPT. Steven Feldman reports on his experiments with mapping LLMs. The result is available here. [h/t quantum of sollazzo]

Caution.

https://www.bespacific.com/like-lawyers-in-pompeii-is-legal-ignoring-the-coming-ai-infrastructure-crisis-parts-i-2-2/

Like Lawyers In Pompeii: Is Legal Ignoring The Coming AI Infrastructure Crisis? (Parts I & 2)

Via LLRX – Like Lawyers In Pompeii: Is Legal Ignoring The Coming AI Infrastructure Crisis? (Part I) – Stephen Embry and Melissa Rogo Rogozinski identify the multiple risk factors involved in the increasing usage of AI in the legal sector, including infrastructure gaps between chip capacity, demand for energy sources and building new data centers, as well as vendor dependencies, promises and deliverables.

Like Lawyers In Pompeii: Is Legal Ignoring The Coming AI Cost Crisis? (Part II) – Stephen Embry and Melissa Rogo Rogozinski challenge the assumption fueling the explosion of AI use in legal is that it will save gobs of time. These savings will inure to the benefit of lawyers and clients, will lead to fairer methods of billing like alternative fee structures, will get better results, improve access to justice, and lead to ‘world peace’. Well, maybe even the vendors would not go so far as to guarantee the last one. But vendors do seem to be guaranteeing everything but that.  And pundits talk as if AI will transform legal from the ground up.  Law firms are buying into the hype, investing in expensive systems that do things they barely understand.

A new slippery slope?

https://apnews.com/article/trump-venezuela-greenland-cuba-571aac35e259857fd512c46f5af11e4d

After Maduro, who’s next? Trump spurs speculation about his plans for Greenland, Cuba and Colombia

A day after the audacious U.S. military operation in Venezuela, President Donald Trump on Sunday renewed his calls for an American takeover of the Danish territory of Greenland for the sake of U.S. security interests and threatened military action on Colombia for facilitating the global sale of cocaine, while his top diplomat declared the communist government in Cuba is “in a lot of trouble.”

The comments from Trump and Secretary of State Marco Rubio after the ouster of Venezuela’s Nicolás Maduro underscore that the U.S. administration is serious about taking a more expansive role in the Western Hemisphere.

With thinly veiled threats, Trump is rattling hemispheric friends and foes alike, spurring a pointed question around the globe: Who’s next?

“It’s so strategic right now. Greenland is covered with Russian and Chinese ships all over the place,” Trump told reporters as he flew back to Washington from his home in Florida. “We need Greenland from the standpoint of national security, and Denmark is not going to be able to do it.”

It’s complicated...

https://philosophy.tabrizu.ac.ir/article_21014.html?lang=en

Knowledge without a Subject: A Philosophical Reflection on the Epistemic Legitimacy of the Machine

The transformation of cognitive and ethical structures in the age of artificial intelligence confronts philosophy with a fundamental question: how does knowledge emerge, and where does responsibility reside when decision-making moves beyond the sphere of human consciousness into algorithmic networks? Focusing on the concept of subjectless knowledge, this study argues that intelligent systems have shifted knowledge from a mental capacity to a mediating process in which human agents, data, and algorithms jointly participate in the production of meaning. Employing a reflective-analytical methodology and examining cases in medicine, media, and law, the paper contends that epistemic validity in the digital era is no longer measured solely by truth, but by the transparency of processes, the capacity for explanation, and the possibility of accountability. Within this framework, social epistemology elucidates how belief is shaped in algorithmic environments, while the ethics of responsibility provides a structure through which the contribution of both human and machinic agents to outcomes can be traced. The proposed model integrates these two dimensions, conceiving knowledge as a mediating event where meaning arises through the interaction between human interpretation and computational reasoning. Accordingly, moral responsibility becomes a distributed property of a network in which every agent participates in the unfolding of cognition. This analysis suggests that maintaining epistemic and ethical legitimacy in intelligent systems requires a philosophical reorientation-from the individual subject toward the distributed architectures of knowing.

Self-driving rules developed by Mad Max?

https://www.mdpi.com/2673-7590/6/1/5

According to Whose Morals? The Decision-Making Algorithms of Self-Driving Cars and the Limits of the Law

The emergence of self-driving vehicles raises not only technological challenges, but also profound moral and legal challenges, especially when the decisions made by these vehicles can affect human lives. The aim of this study is to examine the moral and legal dimensions of algorithmic decision-making and their codifiability, approaching the issue from the perspective of the classic trolley dilemma and the principle of double effect. Using a normative-analytical method, it explores the moral models behind decision-making algorithms, the possibilities and limitations of legal regulation, and the technological and ethical dilemmas of artificial intelligence development. One of the main theses of the study is that in the case of self-driving cars, the programming of moral decisions is not merely a theoretical problem, but also a question requiring legal and social legitimacy. The analysis concludes that, given the nature of this borderline area between law and ethics, it is not always possible to avoid such dilemmas, and therefore it is necessary to develop a public, collective, principle-based normative framework that establishes the social acceptability of algorithmic decision-making.

What strategy? (We can, therefore we must.)

https://www.businessinsider.com/top-general-details-us-military-raid-that-captured-venezuelas-maduro-2026-1

What the top US general revealed about how the surprise 'Absolute Resolve' raid to capture Maduro unfolded in Venezuela

… Dan Caine, the chairman of the Joint Chiefs of Staff, said that the "audacious" mission to extract Maduro — called "Absolute Resolve" — required months of meticulous planning and rehearsal, and involved forces from across the US military.

(Related)

https://www.businessinsider.com/economists-foreign-policy-experts-react-donald-trump-raid-venezuela-maduro-2026-1

Here's what the smartest people in foreign policy, business, and economics are saying about Trump's raid on Venezuela

President Donald Trump on Saturday announced that the US had conducted a raid on Venezuela, resulting in the capture of Venezuelan President Nicolás Maduro and his wife, and big names in business and foreign policy have been reacting as the aftermath unfolds.

Here's what they've been saying:

… Bremmer, founder of the political risk research and consulting firm, Eurasia Group, in a post on LinkedIn, wrote that the "US presumption is next Venezuelan leaders will now do what the Americans want because they've just seen the 'or else.'"