Spyderco Forums

Looks like we've got 2 conversations going on about AI? I would like to be educated on the mechanics of using AI for posts. Maybe a thread dedicated to AI mechanics? How does it work? How can we tell? Maybe it's just me, but I would feel deceived speaking with a bot. Like the movie "her". Who is behind the curtain?

sal

sal wrote: ↑

Fri Mar 24, 2023 3:49 pm

Looks like we've got 2 conversations going on about AI? I would like to be educated on the mechanics of using AI for posts. Maybe a thread dedicated to AI mechanics? How does it work? How can we tell? Maybe it's just me, but I would feel deceived speaking with a bot. Like the movie "her". Who is behind the curtain?

sal

Yea, that's my bad. I saw that bot-ish post over in the Mule forum and cross posted. Probably should have kept the conversation here. I don't know mechanics or how bot posting works. I have not watched "Her," but I did accidentally watch half of "The Crying Game" and I think that evokes a similar reaction. As far as "how can we tell," a few hunches:

* Bot answers seem to be pedantic/didactic/school-marmy
* Often appear to open with a broad, possibly overly-broad, general statement and then narrow down, the classic (but dull and pro-forma) "hour glass" style taught in intro composition classes.
* Bot answers seem to summarize conventional wisdom (and are thus rather dull to read, if you already have a grasp of the facts).
* I've seen the end of a number of bot statements go non-committal and "safe" at the end.
* I've not seen a bot answer that incorporates humor or wit.
* Haven't seen a good bot answer incorporating snark or sarcasm.
* Tend to be responses neither short nor long. Wartstein, for example, definitely isn't a bot. Neither is Beggar-So.
* Very few abbreviations, bad grammar, misspellings, or colloquialisms. Clearly, apollo isn't a bot.
* Middle-of-the-road vocabulary. Few unusual or low-frequency words used.
* You get the feeling you are reading "averages." Ie, they tend to be "safe" answers, unless political, in which case I've not seen a bot answer break right, but have seen many break left.

^ Hope I'm not too OT here. Just my opinions. Not an expert. Interested to see what others think.

I read around a little and found:

“If you have enough text, a really easy cue is the word ‘the’ occurs too many times”

“they are more likely to use common words like “the,” “it,” or “is” instead of wonky, rare words.”

AI text “looked better and contained fewer mistakes. In reality, human-written text is riddled with typos and is incredibly variable, incorporating different styles and slang, while “language models very, very rarely make typos.” (Yeah? Well how long before AI programmers insert the occasional typo?)

Various AI detectors exist. Here’s a web freebie: writer.com/ai-content-detector/

“People tend to overestimate what can happen in one year but underestimate what can happen in 10 years”

Bill Gates

https://www.dailystar.co.uk/news/world- ... s-29535763

We have seen how algorithms and AI have disrupted and impacted society already. Through social media and search engines we get fed some things and it gives us more of what we want further creating divides. There is an argument that he who controls the algorithm and AI controls the future. You don’t have to look hard to see examples. It’s not far fetched to see the AI and algorithms leading the present and future. We have children talking to SIRI and Alexa like they are real people or a kind of Santa.

Can we add spyderco foil lined baseball caps to the merch line up?

Hi Fireman,

We'll begin tooling u on Monday.

sal

sal wrote: ↑

Sat Mar 25, 2023 1:17 pm

Hi Fireman,

We'll begin tooling u on Monday.

sal

Maybe use the powdered metal from the blade profile sanding to add to the aluminum foil that makes us humans less prone to AI. I will be happier knowing some Maxamet and S90V will be adding to my protection

Fireman wrote: ↑

Sat Mar 25, 2023 12:23 pm

Can we add spyderco foil lined baseball caps to the merch line up?

Not to add fuel to the fire but Meta developed this:

https://kotaku.com/facebook-meta-ai-art ... 1849815368

Such a thing if refined further is nothing more or less than an information weapon, that if deployed en mass could destabilize or destroy organizations and communities.