Found on my travels... Phind

Café
#21

I have used this and gave it a tough query

first 8 characters tiddlywiki filter syntax

It didn’t find a solution. But then I did some more sleuthing around and searched on DuckDuckGo

first 100 characters tiddlywiki

Which led me to Grok TiddlyWiki which eventually led me to this link. And I ended up using as my final solution

[{!!title}split[]first[8]join[]]

if you then put in “first 100 characters tiddlywiki” into Phind, it also shows Grok TiddlyWiki as a source. You can then hover over the source and choose “Regenerate only with domain” and the results are pretty impressive.

I see many of the conversations around AI being quite dismissive, mostly by “regular people” who are not developers who just are repeating whatever they are hearing from the news based on their speculation about the technology rather if they had actually used it. Of course we cannot expect AI will replace developers. This is part of the problem with the terminology is it creates a disconnect between what people think AI is and what it actually is.

I will probably keep using this because it provides another interesting way to get unstuck when you are searching for a solution to a problem.

1 Like
#22

AI is now illustrating the uncanny valley between plagiarism and what we have assumed to be the entirely distinct domain of original thought.

Of course it’s actually a multi-dimensional continuum. Learner-level participants in a complex field must start out by mimicking others’ patterns (plagiarism, copying, rehearsing), and gradually get the hang of inhabiting the pattern-space fluently. When I speak Japanese, I still “plagiarize” quite a bit – treating someone else’s speech as already-competent, and tweaking it only slightly to fit the current situation. :wink:

AI’s path toward verbal fluency looks a bit different of course. It’s clunky in places human speakers tend not to be; yet it synthesizes a quantity of available discursive resources that we cannot absorb.

I had my first case (as far as I know!) of AI-written work submitted by a university student, recently. I was able to sniff it out; the uncanniness was hard to put a neat finger on, but palpable enough. I suspect it will get more difficult very soon!

One lesson is that “plagiarism” vs non-plagiarism was never actually as simple as we could get away with thinking it was.

#23

I current withhold calling this A.I. a subject I have being interested in for a long time, Large Language model LLM suits me.

It seems to me its key value is that it can be used as a well informed assistant. You still have to ask the right questions to a useful end, and trust or always question your assistants capacity.

There are threats from it where on one hand I may no longer need to work in a team, because my LLM is my assistant(s), on the other hand others may not employ me as their assistant.

  • Its early days, and it is revolutionary, but history has taught us to be skeptical and the hype cycle has only started.
  • In another decade we will again ask ourselves why did we misread it, why did we not understand the implications and why did we believe so much b–llsh-it
  • Remember it can be a curse to live in interesting times, but I am glad I do :nerd_face:
#24

honestly, yeah I agree with you here.

I haven’t yet met something that I can classify as true A.I. but I don’t feel like what we have now a days is much more than an LLM.

Wouldn’t an A.I. have the ability to ask it’s own questions and create it’s own concepts through retaining information and by that extent, learning?

food for thought, ofc.

1 Like