Showing posts with label ChatGPT. Show all posts
Showing posts with label ChatGPT. Show all posts

Sunday, October 29, 2023

Technology's Ethical Two-Step

[B&W sketch of a man in a ballroom dance with a robot, dressed in a dress.]

1. Now. Delay incorporation of ethics. Let’s not muddy the waters in a way that holds back Progress.

2. Later. Deny incorporation of ethics. It’s too late. People have come to rely on things as they were built. It would be Disruptive to change now.


If you got value from this post, please “Share” it.

I've said things vaguely like this for a long time, but I packaged it up crisply like this in a post on Mastodon, for which this is a mirror.

Original Keywords were described as: “Ethics, Tech, Technology, Society. Presently very relevant to, but not exclusive to: AI, ML, LLM, GPT, ChatGPT.”

I made a later edit to this post to add a graphic, ironically generated by Abacus.AI's GPT-4o ChatLLM chatbot calling out to FLUX.1. The prompt was "make me a black and white image that depicts sketch of two entities engaged in a ballroom dance, one a man and his partner a robot."

Friday, July 14, 2023

Lying to Ourselves

My friend David Levitt posted this hypothesis on Facebook:

Theory:
Humans are so mentally lazy and emotionally
dishonest about what they know, soon AI will
be much better leaders.

I responded as follows. Approximately. By which I mean I've done some light editing. (Does that mean I lied when I say this is how I responded?)


I think the notion of honesty here is a red herring. There are a lot of human behaviors that do actually serve a purpose and if you're looking for intellectual honesty, it's as much missing in how we conventionally summarize our society as in how we administer it or ourselves.

Of course we lie sometimes.

  • We lie because not all answers are possible to obtain.
    What is an approximation to pi but a lie?
  • We lie because it comforts children who are scared.
  • We lie because it's more likely to cause success when you tell people your company is going to succeed than if you say "well, maybe" in your pitch to rally excitement.
  • We lie because it saves face for people who tried very hard or never had a realistic chance of affecting things to tell them they are blameless.
  • We lie because some things are multiple-choice and don't have the right choice.
  • We lie because it protects people from danger.
  • We lie because some things happen so fast that abstractions like "now" are impossible to hold precise.
  • We lie because we are imprecise computationally and could not compute a correct truth.
  • We lie because not all correct truth is worth the price of finding out.
  • We lie because papering over uninteresting differences is the foundation of abstraction, which has allowed us to reason above mere detail.
  • We lie because—art.

So when we talk of machines being more intellectually honest, we'd better be ready for what happens when all this nuance that society has built up for so long gets run over.

Yes, people lie for bad reasons. Yes, that's bad and important not to do.

But it is naive in the extreme to say that all lies are those bad ones, or that of course computers will do a better job, most especially computers running programs like ChatGPT that have no model whatsoever of what they're doing and that are simply paraphrasing things they've heard, adding structural flourishes and dropping attribution at Olympic rates in order to hide those facts.

Any one of those acts which have bootstrapped ChatGPT, by the way, could be called a lie.


Author‘s Notes:

If you got value from this post, please “share” it.

Laziness is also misunderstood and maligned, but that is topic for another day. For now, I refer the ambitious reader to an old Garfield cartoon that I used to have physically taped to my door at my office, back when offices were physical things one went to.

Sunday, February 12, 2023

Strange Poetry and ChatGPT

I should say as context that I had been having a conversation on Mastodon in which I got the idea to write a poem that was somehow both a haiku and a limerick. I wrote the following myself. (The ChatGPT part comes later.) I've numbered the lines to make it more intelligible:

   0   There once was a would-be haiku
   1   Wedged late in a lim'rick venue
   2   Ending a hero
   3   Numbered from zero
   4   Its lines went three, four and two

If it helps in understanding it, I went on to say in the other forum, the haiku to be extracted and rearranged is this one:

  Numbered from zero
   Its lines went three, four and two
    Ending a hero

But while I'm proud I was able to come up with anything that was both limerick and haiku, I have to admit it's a more boring haiku than I'd aspired to. So I later wondered if ChatGPT could help me do better. That's what led to the following odd exchange a few days ago, in which it's clear that ChatGPT can do amazing things in some ways, yet be hampered with sometimes the simplest of things in others.

I came to wonder through this if it even knows what a syllable is. Perhaps it has just seen so many haikus that it doesn't need to count, and when asked to write something similar to, but not the same as, a 5-7-5 haiku, it doesn't know what the goal is at all?

See what you think...

write a haiku about the beauty of a snowy day

Snowflakes dance down
Peaceful white blanket unfolds
Winter's pure delight

write a limerick in which the last three lines, if rearranged, would be syntactically valid as a haiku and in which the first two lines introduce that haiku.

In the park, a snowflake lands
A symbol of winter's grace
Peaceful, it descends
Covering earth's face with white
Nature's beauty in sight.

that isn't quite right. in a limerick, the first, second, and fifth line are long. so to embed a haiku in a limerick, the embedded haiku will have to use the form 5-5-7.

Note that the apology here is part of the limerick! Is that it being clever or just revealing it doesn't know the difference between the task and related meta-conversation?

Oh, my apologies, I see
Let me write it now correctly
A snowflake takes its flight
Brings peace to the winter night
In a dance of grace, so free.

Write a fresh 5-5-7 haiku, in other words with the long line being last.

Beneath a winter sky
Snowflakes gently fall like dreams
Peaceful and serene.

In a 5-5-7 haiku, the second and third lines are exchanged.

Gently falling snowflakes
Beneath a peaceful winter sky
Serene dreams descend.

No. It should say "Gently falling snowflakes. / Serene dreams descend / Beneath a peaceful winter sky." so that 5-5-7 describe the number of syllables in each line, 5 for line 1, 5 for line 2, and 7 for line 3.

Gently falling snowflakes,
Serene dreams descend,
Beneath a peaceful winter sky.

Yes, now modify the poem so the first and second lines rhyme.

Snowflakes fall, so gentle and light,
Peaceful dreams take flight in the night.
Beneath winter skies so serene,
Nature's beauty, a lovely scene.

But it should still be a 5-5-7 haiku.

Falling snowflakes, so pure,
Peaceful dreams that softly tour,
Beneath winter sky serene.

In a 5-5-7 haiku, there are 5 syllables for line 1, 5 syllables for line 2, and 7 syllables for line 3.

Snowflakes dance, so light and slow
Peaceful dreams, a winter's glow
Beneath the sky, so serene and bright.

That is a 7-7-9 haiku you wrote. Tighten it up into a 5-5-7 haiku by using fewer syllables.

Snowflakes twirl, light and slow
Dreams of peace, a winter's show
Beneath the sky, serene and bright.

Author's Notes:

Thanks to Madeleine Begun Kane for writing a haiku about haikus and limericks and putting it on Mastodon, which inspired me to write my own original poem and to have the enclosed conversation.

It's fascinating out chance events like that influence our lives. My whole career has been littered with chance events that were as important as anything planned. (Programming language enthusists might enjoy my paper Common Lisp: The Untold Story, a record of an invited talk I gave at the 50th Anniversary of Lisp conference, where I recount a number of those. No, not in poetry form. Poetry, and especially a haiku, is a kind of apology or penance that I do to balance the fact that some of my other writings are quite long.)

There is some additional discussion of this blog post and ChatGPT in general where I mentioned it on Mastodon.

You can try ChatGPT here if you want.

All of the "haiku" in here is really senryu.

If you got value from this post, please “share” it.