# Lmf's quick questions about CF

**URL:** https://discuss.criticalfallibilism.com/t/lmfs-quick-questions-about-cf/1906
**Category:** Unbounded
**Created:** [January 17, 2024, 1:56am UTC](https://discuss.criticalfallibilism.com/t/lmfs-quick-questions-about-cf/1906 "2024-01-17T01:56:52Z")
**Posts on this page:** 1
**Showing post:** 8

<div class="post-metadata">

### Author: ![lmf](https://discuss.criticalfallibilism.com/user_avatar/discuss.criticalfallibilism.com/lmf/32/1302_2.png) [@lmf](https://discuss.criticalfallibilism.com/u/lmf)
#### Post date: [January 23, 2024, 1:53am UTC](https://discuss.criticalfallibilism.com/t/lmfs-quick-questions-about-cf/1906/8 "2024-01-23T01:53:04Z")

</div>

In an [old thread](https://discuss.criticalfallibilism.com/t/career-physics-and-goals-was-artificial-general-intelligence-speculations/), ET emphasized the following philosophical principle:

> [@Career, Physics and Goals (was: Artificial General Intelligence Speculations)](https://discuss.criticalfallibilism.com/t/career-physics-and-goals-was-artificial-general-intelligence-speculations/473/128):
>
> If you don’t know the (underlying) cause of an error, then you don’t know the consequences of that (underlying) error, so it’s not safe to ignore it.

**Q2) What is meant by “safe”?**

My normal way of thinking about the safety of ideas is that it is a relative term. Because of fallibility, no conceptual identification is ever _perfectly_ safe, so technically, we can only say that some are _more safe_ or _less safe_ than others. But that contradicts CF, because it is an epistemological degree judgement. As a result, I don’t have a crystal clear idea of what is meant. (Does it mean that the idea that you should ignore the error is refuted? If so, why? What refuted it? And relative to what goal?)

Here’s an example that may further clarify my confusion:

ET, like everyone else, sometimes makes typos. A typo is an error, because when someone types a word, he has the implicit goal of spelling it correctly (with some exceptions, e.g. if the goal is to quote somebody who made a typo).

It is hard for me to imagine that ET actually understands the underlying cause of why he makes typos. Knowing that would require a knowledge of the inner workings of the human brain that I don’t think anyone possesses at present. He probably knows some things about his typos, like I bet he has some knowledge about situations where they are more likely to occur, but that isn’t the same as understanding a cause.

So in what sense is it _unsafe_ for him to ignore typos? He seems to think (rightly) that his typos are not a big deal.

---

_[View the full topic](https://discuss.criticalfallibilism.com/t/lmfs-quick-questions-about-cf/1906)._