The Ghost Has No Intelligence Without Us Being There
The enthusiasm about the potential of artificial intelligence (AI) seems to miss something, an obvious question, about the actual quality and character of AI-generated content and conversations.
I have to think about this and express the actual problem accurately. The problem is less with AI than apparently with our ability to differentiate the qualities and character of our own conversations.
A conversation with a purpose as simple as, Pass the salt, and the consequent actions, is at once very human and also unambiguous, or so it seems.
Even if I do not add more to the description, you imagine the rest. You might add ambiguities. You are capable of making something that simple into so much more. AI cannot really participate in that interaction.
A conversation about anything requiring action in the sense of humans doing something beyond talking is one-sided if a bot is the only other participant.
This includes further thinking as well as physical action.
The bot is not going to develop a body and jump out of its nowhereness to dance with its Pygmalion no matter how excited a programmer gets about it.
Likewise, the chatbot does not learn or pursue further ideas in anything like the way a human does.
The fault is not with AI but with our consideration, our thinking about our own thinking and the neat parlor tricks computers and programs are capable of producing.
Take the deep fake on the website The Infinite Conversation.
AI here is being used to create conversation between Herzog the filmmaker and Slovenian philosopher Slavoj Žižek (whom I’d never heard of or from, he doesn’t call, he doesn’t write).
The quality of this conversation is strained to say the least. Not only that, it genuinely, authentically, does not go anywhere in a discernible human sense.
Neither of the bots convinces the other of anything. They are not even exchanging ideas or responding in anything but the most superficial senses to each other.
We could program AI to make puns and associate, after the fashion of literary banter perhaps. But what would be missing, at least at present, is the nuanced sense that humans bring to actual conversations most of the time.
Laughter and groans might be possible but would they convey the same meanings and feelings to the humans, and lead to the follow-ups humans engage in without thinking?
I have a long history, well before technology reached the point it has now, of discussing with friends the kinds of exchanges we humans have with each other.
As one example, someone tells a story about having fallen off a bicycle. The listeners are reminded of similar stories in their own experiences, and the conversation takes the form of shared anecdotes.
That can easily shift to another kind of conversation, one of shared advice or expressed admiration. Questions interject themselves about the kind of bicycle and perhaps innovations in technology of bikes generally.
Think back on the example of asking for the salt. That kind of exchange pops up in the middle of the other kinds of exchanges.
To make it simple, consider the scripting of an operatic production impromptu and sung in perfect harmony and expert timing like everybody kung fu fighting fast as lightning a little bit frightening, and the next thing the little boy who was sitting quietly the whole time is saying good-bye as you breathe your last breath, years later as you both recall the previous conversation.
In short, we seem to be tone deaf and unconscious to the sophistication of the composition and musical qualities of even our most mundane interactions over a lifetime.
Something else is going on. When we imagine AI bots are doing anything that approaches what we do mostly without thinking, it is only because we are there imagining both our side and another side that is just not there without us.
Like
Comment
Share
I Dwell In Possibility 