Recognition is growing that fundamental challenges make LLMs unreliable. Increasingly expensive scaling will likely hasten the popping of the AI bubble.
Read More ›
AI models are really good at finding patterns in data but they are bad at deciding whether these spotted patterns can be used to make reliable predictions.
Read More ›
With the Yankees and Dodgers so evenly matched, this World Series will be somewhat like a coin-flipping contest. That’s the paradox of luck and skill.
Read More ›
Lichtman’s Thirteen Keys to the White House prediction model for the presidency shows some evidence of p-hacking, as does Helmut Norpoth’s Primary Model.
Read More ›
Some prediction models were developed by well-intentioned researchers before the perils of p-hacking were clearly understood, hence the failures.
Read More ›
It continues to astonish us how easily people can be persuaded that LLMs (chatbots) can do our thinking for us
Read More ›
For example, the baseball player with the highest batting average in any season generally does not do as well the season before or the season after.
Read More ›
Statistical analysis shows that in major sports, the players or teams that won most of their games benefitted to a surprising degree from good luck.
Read More ›
Reinhart/Rogoff’s dismissal of criticism of their deeply flawed study as “academic kerfuffle” is unconscionably cavalier.
Read More ›
The current, widely-favored 60/40 strategy has little or nothing to recommend it beyond the fact that it is what everyone else is doing.
Read More ›
Not knowing what words mean, neither OpenAI’s ChatGPT 3.5 nor Microsoft’s Copilot nor Google’s Gemini can do a simple logic test.
Read More ›
It takes an experienced financial planner to distinguish between good and bad advice, so clients may as well skip the LLMs and go to the knowledgeable human.
Read More ›
Not understanding what words mean or how they relate to the real world, chatbots have no way of determining whether their responses are sensible, let alone true
Read More ›
As an economist, professor, and writer, I am reportedly in imminent danger of being replaced by large language models (LLMs). I’m not worried.
Read More ›
No matter how many words they train on and no matter how many human trainers correct their mistakes, they will continue to lack the critical thinking skills.
Read More ›
Based on extensive training on untold amounts of text, LLMs are able to repackage superficially compelling answers that they literally do not understand.
Read More ›
Decades of geniuses trying to build computers that are as intelligent as they are have shown how truly remarkable our brains are—and how little we understand.
Read More ›
Internet-based businesses tend to follow a life cycle in which quality deteriorates over time. Writer Cory Doctorow calls the process “enshittification.”
Read More ›
The hallucinations are symptomatic of generative AI models’ core problem: they can’t identify output problems because they know nothing about the real world.
Read More ›
In my book Distrust (Oxford 2023), I recommend that journals not publish data-driven research without public access to nonconfidential data and methods used.
Read More ›