The question to ask ...

How convincing does the illusion of understanding have to be before you stop calling it an illusion?
Animations by Zach Lieberman

Wish this awesome New Yorker animation was a gif and not a video as it's truly stellar to a fault but this blurb is not about that, it's about the illusion of thinking as expressed by AI as a frenzy of talk regarding AGI and ASI is becoming ever more animated as we speak.

Dario Amodei, the C.E.O. of the artificial-intelligence company Anthropic, has been predicting that an A.I. “smarter than a Nobel Prize winner” in such fields as biology, math, engineering, and writing might come online by 2027. He envisions millions of copies of a model whirring away, each conducting its own research: a “country of geniuses in a datacenter.” In June, Sam Altman, of OpenAI, wrote that the industry was on the cusp of building “digital superintelligence.” “The 2030s are likely going to be wildly different from any time that has come before,” he asserted. Meanwhile, the A.I. tools that most people currently interact with on a day-to-day basis are reminiscent of Clippy, the onetime Microsoft Office “assistant” that was actually more of a gadfly. A Zoom A.I. tool suggests that you ask it “What are some meeting icebreakers?” or instruct it to “Write a short message to share gratitude.” Siri is good at setting reminders but not much else. A friend of mine saw a button in Gmail that said “Thank and tell anecdote.” When he clicked it, Google’s A.I. invented a funny story about a trip to Turkey that he never took.

The question to ask ...

I once had a boss who said that a job interview should probe for strengths, not for the absence of weaknesses. Large language models have many weaknesses: they famously hallucinate reasonable-sounding falsehoods; they can be servile even when you’re wrong; they are fooled by simple puzzles. But I remember a time when the obvious strengths of today’s A.I. models—fluency, fluidity, an ability to “get” what someone is talking about—were considered holy grails. When you experience these strengths firsthand, you wonder: How convincing does the illusion of understanding have to be before you stop calling it an illusion?

A leading neuroscientist argues that ChatGPT “radically demystifies thinking.”

The question to ask/rev II

Artificial neural networks compress experience just like real neural networks do. One of the best open-source A.I. models, DeepSeek, is capable of writing novels, suggesting medical diagnoses, and sounding like a native speaker in dozens of languages. It was trained using next-token prediction on many terabytes of data. But when you download the model it is one six-hundredth of that. A distillation of the internet, compressed to fit on your laptop. Ted Chiang was right to call an early version of ChatGPT a blurry JPEG of the web—but, in my view, this is the very reason these models have become increasingly intelligent. Chiang noted in his piece that, to compress a text file filled with millions of examples of arithmetic, you wouldn’t create a zip file. You’d write a calculator program. “The greatest degree of compression can be achieved by understanding the text,” he wrote. Perhaps L.L.M.s are starting to do that.

One never knows, do one? Fats Waller