The Dangerous Economist: What if Alan Turing and Claude Shannon discussed the state of artificial intelligence right now?

Thursday, April 23, 2026

What if Alan Turing and Claude Shannon discussed the state of artificial intelligence right now?

Here is a prompt I gave ChatGPT:

"You are a journalist living in the year 2026 who specializes in science and technology. You are also an historian of science and technology. Write a dialogue that might occur between Alan Turing and Claude Shannon on the state of artificial intelligence right now in the year 2026. Include what they think are the current good and bad points of AI right now as well as what they think the next big improvements will be"

Here is the result:

"A Fictional Dialogue in 2026: Alan Turing Meets Claude Shannon

Setting: A quiet study filled with books, chalkboards, and a faint hum of modern servers in the background. Alan Turing and Claude Shannon, somehow transported into 2026, sit across from each other, each examining a tablet displaying the latest AI systems.

Turing:
It seems my old question—“Can machines think?”—has been… not answered, but thoroughly complicated.

Shannon:
That’s a polite way of putting it. I’d say machines now simulate thinking with remarkable efficiency. They process language, images, even emotions—at least statistically.

Turing:
Yes, I’ve seen these large language models. They can converse fluently, generate essays, even write code. In some sense, they pass what people now call the Turing Test rather routinely.

Shannon (smirking):
You’ve accidentally made yourself obsolete, Alan.

Turing:
Or perhaps the test was too lenient. These systems don’t understand in the way humans do. They predict. They approximate. They lack grounding in experience.

Shannon:
That’s because they’re built on probability distributions—my territory. These models compress vast amounts of human knowledge into statistical patterns. Elegant, really. But brittle.

Turing:
Brittle indeed. I’ve noticed they can be confidently wrong. Hallucinations, they call them. Quite troubling.

Shannon:
Noise in the channel. When the signal isn’t strong enough, the system fills in the gaps. The difference is, people trust the output far more than they should.

On the Strengths of AI in 2026

Turing:
Still, we must acknowledge their strengths. Their ability to assist in scientific discovery is extraordinary. Drug design, protein folding, climate modeling—tasks that would take humans decades.

Shannon:
And communication! These systems translate languages instantly, summarize entire libraries, and act as interfaces between humans and machines. Information theory has found its ultimate playground.

Turing:
They also democratize expertise. A student in a remote village can now access something resembling a personal tutor.

Shannon:
Yes, though whether that tutor is always correct is another matter.

On the Weaknesses and Risks

Turing:
I worry about their lack of reasoning. They appear intelligent, yet fail at basic logic in unexpected ways.

Shannon:
They optimize for plausibility, not truth. That’s a dangerous distinction.

Turing:
And then there is the social impact—misinformation, automated propaganda, deepfakes. Machines amplifying human biases at scale.

Shannon:
Not to mention economic disruption. Automation is replacing not just manual labor, but cognitive work—writing, design, even programming.

Turing:
A paradox: machines that assist us, yet threaten to displace us.

Shannon:
That’s always been the story of technology. The difference now is speed. The feedback loop is accelerating.

On the Nature of Intelligence

Turing:
Do you think these systems are truly intelligent?

Shannon:
They are information processors of unprecedented scale. Intelligence? That depends on your definition. They lack goals, intentions, awareness.

Turing:
Then perhaps intelligence is not a single property, but a collection of capabilities—many of which these systems now possess.

Shannon:
You’re redefining the term to include them.

Turing:
Or expanding it.

On the Future: What Comes Next

Shannon:
If I had to guess, the next leap will be in reasoning. Systems that don’t just predict words, but can follow chains of logic reliably.

Turing:
Yes—models that can plan, verify, and correct themselves. Less like parrots, more like mathematicians.

Shannon:
And grounding. Connecting these models to the physical world—through sensors, robotics, real-time data.

Turing:
Embodiment. A machine that experiences the world may finally begin to understand it.

Shannon:
There’s also efficiency. These current systems are absurdly resource-intensive. The future will demand smarter, leaner architectures.

Turing:
And alignment. Ensuring these machines act in accordance with human values.

Shannon (raising an eyebrow):
Whose values?

Turing:
Precisely the problem.

Closing Reflections

Shannon:
It’s fascinating. We built machines to process information, and now they process us—our language, our thoughts, our culture.

Turing:
And in doing so, they hold up a mirror. Perhaps what we are really studying… is ourselves.

Shannon:
So, Alan—have machines begun to think?

Turing (after a pause):
They have begun to convince us that they do. The rest remains… an open question.

The servers hum louder, as if listening."

Click here to read the Wikipedia entry on Turing. He lived from 1912-1954. Excerpt:

"English mathematician, computer scientist, logician, cryptanalyst, philosopher and theoretical biologist. He was highly influential in the development of theoretical computer science, providing a formalisation of the concepts of algorithm and computation with the Turing machine, which can be considered a model of a general-purpose computer. Turing is widely considered to be the father of theoretical computer science."

Click here to read the Wikipedia entry on Claude Shannon. He lived from 1916-2001. The AI program Claude is named after him. Excerpt:

"American polymath who was a mathematician, electrical engineer, computer scientist, cryptographer, and inventor known as the "father of information theory" and the man who laid the foundations of the Information Age."