Two items which are compatible with previous discussions here:
- Eurisko is guilty of dialogue with the human
- Invariants change
Let’s take it one by one.
Eurisko is guilty of dialogue with the human. The source is this, the video is this. During the video of the demo, the Eurisko code is run with void results. Why? Apparently because back then the machine was much more slower than the virtual machine used now and it had much less RAM. Therefore, back then, the human (Doug Lenat) who created the AI Eurisko, had enough time to read all the output of Eurisko, to think about which part is interesting and to add the interesting part in the Agenda. Guilty of dialogue! Today, even if the AI is trained on vastly bigger data, we don’t talk with it. You would say we do, all these chats, but no! the AI does not get retrained on items the user considers interesting… there is no collaboration, there is no dialogue. Discussed many times, oldest probably in the Morlocks and eloi post, quote:
“From historical reasons maybe the morlocks (technical nerds) are trained/encouraged/selected to hate discussions, human exchanges and interactions in general. Their dream technology is one like in (1), i.e. one which does not talk with the humans, but quietly optimize (from the morlock pov) the eloi environment.”
Invariants change. Source is The problem with invariants is that they change over time. Quote:
“
- Our systems change over time. In particular, we will always make modifications to support new functionality that we could not have foreseen earlier in the lifecycle of the system.
- Our code often rests on a number of invariants, properties that are currently true of our system and that we assume will always be true.
- These invariants are implicit: the assumptions themselves are not explicitly represented in the source code. That means there’s no easy way to, say, mechanically extract them via static analysis.
- A change can happen that violates an assumed invariant can be arbitrary far away from code that depends on the invariant to function properly.
What this means is that these kinds of failure modes are inevitable.”
What is the problem here? Well if you think we live in the deterministic, symmetric and semantic universe from Which way true?, then of course that you are a hypocrite who knows well that this can’t be true, but you have to make believe it is. Quote:
“There has to be a formal mistake in this formal game. Despite all claims that the game is well written and everything is top done, there is a secret hope that hacks are possible because the devs were sloppy somewhere. And why not? After all, by symmetry we learn all the time that any particular piece of the universe which looked varied, it is not. Our uniqueness it is not. Why? Because somebody says so? In the simple symmetric and explainable universe everybody knows that the rules are imposed, not natural. “
chorasimilarity 