Unmaking Sense

Episode 8.30: Thoughts and thinking in ChatGPT and you and me.

If the processes that go on in our neurological infrastructure are unintelligible, and I think there are a great many reasons to believe that they are, and if the same is true of what goes on in a neural net to beneath the surface of its output, then that again is another parallel between the way chatbots, or this particular chatbot chatGPT and human beings operate. That we cannot interrogate these inscrutable infrastructures to check up on them should make us more cautious about trusting their output m, just as we are cautious about trusting the output from human beings. But that chatGPT is no more considered to be infallible than you or me may in the end prove to be one of its most endearing characteristics, for if it were ever to be regarded as a universal oracle incapable of error, the world would be in much more danger than it is when we treat it as something that we have to take and scrutinise and challenge and question and analyse before we act too trustingly upon what it tells us. In that respect it mirrors human interactions.

Episode 8.29: ChatGPT and I share this: the first time we say something may be the first we think it

Sometimes we prepare things in advance by writing them down, discussing them with other people, drafting them re-drafting them, editing them and finally we let them loose upon the world. I suspect that when chatGPT responds to a user’s question, it is probably the first time that that particular concatenation of words and sentences has ever arisen for it or in the world. The way I record these podcasts is just the same, for good or ill.

Episode 8.28: Memory and personal identity; implications for AGI and chatGPT.

What role does memory play in defining and accessing self-awareness? Compressibility and human lives; discovering the irreducible. Discovering things that are incompressible, so cannot be summarised and must be experienced in real time in order to be understood.

Episode 8.27: Because memory is the chief source of personal identity, a chatGPT can be self-aware.

We dedicate the episode to revisiting David Hume‘s brilliant description of the nature of personal identity in terms of the association of ideas and our memories of things, and we draw the conclusion that we’re we to facilitate and require a chat bot to store and be aware of and take account of the memories of its own output as a part of its contemporary input, that would overcome a principle obstacle to it becoming self-aware.

Episode 8.26: Episodic alternatives to consistency.

There is no particular reason why we should be consistent with our own behaviour in the past, whether with what we’ve said, or what we’ve done, or what we’ve thought, or what we’ve written. It’s perfectly possible - there is no moral or universal law forbidding it - to be inconsistent, to make every action that we perform independent of all those that we have taken in the past, and perhaps of those that we will take in the future. This is called being episodic, or living an episodic rather than a narrative life. It’s an interesting idea, and this little episode is about it. The more episodic a life becomes, the less compressible it may be.

Episode 8.25: What is sentience or self-awareness? Why a chatbot like chatGPT could be self-aware.

We consider what we mean when we talk about something being self aware, and we see that there’s nothing very remarkable about it at all other than an ability to use one’s own previous output as new input, and therefore something in the world with which one can have a conversation. On this basis there is no reason why a chat bot like ChatGPT with some small software tweaks could not become self-aware.

Episode 8.24: All life including chatGPT conceived as irruptions from the depths of the universe.

We use the analogy of bubbles rising to the surface of a mud pool and appearing as blips on the surface as a way of illustrating the way in which all life should be conceived as elements of the universe in its striving to relate to and scrutinise itself. We consider some implications for the way we make sense of those irruptions, and why drawing connections between surface appearances including ourselves, our lives and actions and words, is not the right approach to understanding them.

Episode 8.23: ChatGPT is the latest way in which the universe has contrived to scrutinise itself.

We try to carry out the kind of reconfiguration of our outlook that we described in the previous episode, and in so doing obtain a new perspective on the significance or otherwise of the chatGPT phenomenon and our response to it. With apologies for the poor audio quality.

Episode 8.22: ChatGPT is just the latest in an evolutionary sequence, and it will not be the last.

We consider chatGPT as just the most recent step on an evolutionary journey which began long before human beings appeared upon the scene, and will no doubt continue long after they leave it.

Episode 8.21: Hacking ChatGPT to make it do your bidding is much harder than you might think.

A natural concern about any artificial general intelligence is that an android exhibiting it might somehow be hacked by some external malignant force. We consider how plausible this is given the particular way in which a neural net like ChatGPT is trained and operates, and come to the conclusion that it is virtually impossible. We draw some analogies with the way human beings use their brains and learn.