With all the talk of general AI taking over the world, a short rant.
Language models will never become smart. In the same way that the book you read most recently will never itself become smart.
They are a very good way to deal with large amounts of data and synthesise an output from it. That is the whole of it. There is not one shred of intelligence in there.
What would actually be required
At best, language models may one day be a tool used to educate a true general AI. But until there are sufficiently large numbers of reliable quantum nodes, that is not happening, and we are barely at the abacus stage there.
For scale, at the time of writing the largest array of qubits assembled was around a thousand. The human brain has hundreds of billions of interconnections.
The worry that is actually warranted
If there is something to worry about with language models, it is that, like any tool, humans may use them for bad things. Nothing more exotic than that.
Which is a more useful thing to plan around than the science fiction version, because it is the one that actually turns up.
Itwerx Corp is a service-disabled veteran-owned small business providing IT services across Seattle, Bellevue, Everett and Snohomish County. This is the kind of thing our AI integration work deals with – talk to us about yours.
