Importantly, the group avoids blaming people for errors. 
  The process assumes blame - and it's the process that is 
  analyzed to discover why and how an error got through. 

  At the same time, accountability is a team concept: 
  no one person is ever solely responsible for writing or 
  inspecting code. "You don't get punished for making errors," 
  says Marjorie Seiter, a senior member of the technical staff. 
  "If I make a mistake, and others reviewed my work, then I'm
  not alone. I'm not being blamed for this."

I admire your breakdown, and I felt your pain.

My frustration was with opening a new tab for each sample because I was too lazy to transcribe rolls (for review) -- so once I'd opened and revisited about 16 tabs I got rather annoyed and just started rapid rolling looking for high values ...

After about 30 seconds of this pique I rolled a Yahtzee 55555 value 20 and went: Doh!

My takeaway is that using "induction" to illuminate Human ~vs LLM may be wrong framing when Eureka moments are possible.

Matrix Reloaded (2003) ...

  Neo: Hiya fellas.
  Agent 1: It’s him.
  Agent 2: The Anomaly.
  Agent 3: Do we proceed?
  Agent 2: Yes.
  Agent 3: He is still...
  Agent 1: ...only human.
  Neo: Hmm--*upgrades* [beats all three nonetheless]

  The shareholders won't be happy.

  It's not our job to make them happy. 
  It's our job to make them money, 
  but once we launch this little baby, they're gonna flip!

  We call this Pure O2. 
  This is the first of our planned upgrades. 
  Once we can roll back some of Halliday's ad restrictions, 
  we estimate we can sell up to 
  80% of an individual's visual field 
  before inducing seizures.

The dimensionality of color vision in carriers of anomalous trichromacy (2010)

https://jov.arvojournals.org/article.aspx?articleid=2191517

So I suppose we could look forward to genetic enhancement in order to perceive better display technologies

or perhaps we are far enough along this timeline that the AI editor preferred the distinction for the story title?

^ = from your linked Toyota Production System philosophy

Thanks for that link! I find those two early sentences to be an insightful and relatively complete loop of process.

When considering using automation or ~A.I. we can easily ask: which part of this loop is our addition improving (or messing up)? Where is the balance that works?

As you point out, answering these what-works/quality questions are not solved problems and expertise is needful, but careful consideration does not seem to be a popular mode in our age of fountaining funny money and magic genies.

I grew up in the '60s where the science fiction/future was always: march of progress and you'll have so much time and choices! Now I am in the future and the reality is close to: who has the time?

More insidiously/invisibly: you have plenty of time for endless momentary thrills, but no one has time to make things better.

Once upon a time there were customer complaint departments and the Production System would get fixed. Then it became suggestion boxes and returns. Then pleading for a Return Merchandise Authorization. Then it's your call is valuable to us recordings before click call hangups. Oops, unhappy customer?--give 'em a coupon to keep up the addiction, it's cheaper than Quality Control.

The latest devolution seems like fire the workers but use AI to mesmerize customers, or just mind control ~investors and ~regulators, since who needs cover-our-costs-paying-customers anyway?

Will the pendulum swing back?

  I find that when someone's taking time to do something right in the present, 
  they're a perfectionist with no ability to prioritize, 
  whereas when someone took time to do something right in the past, 
  they're a master artisan of great foresight.

In OP's terminology, I would add that the left-path ascenders are playing king-of-the-hill as they know that (in our current business environment) they can be the one to roll down rocks from the top to both crush those latecomers attempting to overtake them, as well as cheaply flatten paths up for the consumer masses.

Those masses will pay this King for an easy ascent because, above all, the Hormuz spice must flow.

Who has time to make their own donuts?!

From this kind of P.O.V. the OP's middle and right paths are often just losing strategies whose appeal lies in their comfort & morality (i.e. a luxury of personal preference).

To be clear, I am not personally advocating for Capitalist Utilitarianism here, but I find it worthwhile to see our current turmoil from this higher viewpoint and perhaps relevant to some personal software development decisions.

The Moon is a harsh Mistress.

https://daringfireball.net/preferences/

It seems to me that copyright doesn't exist anymore as a reasonable notion and effective right for an individual creator--A.I. being the final stake through the heart

Comments URL: https://news.ycombinator.com/item?id=48202351

Points: 5

# Comments: 2

> The signal you want to send isn’t “I spent effort on this,” because that signal is now cheap to fake. The signal you want to send is “a human actually wrote/made this.” That one currently costs almost nothing to send, because so few people are sending it.

or, in other words ... "when the going gets weird, the weird turn pro" -- From Fear and Loathing in Las Vegas (1971) by Hunter S. Thompson

P.S. so, was this, elegantly composed, brilliantly thought provoking, comment sufficiently Weird enough?

or am i ai ?

Re-reading this short essay (after many years) got me to wondering if those of you using AI coding assistance tools are using them to both analyze and document existing code (i.e. non AI gen'd code but AI gen'd coments/doc for humans to read) or only transiently (e.g. for your own quick understanding or to help the AI whip up an API etc)

I also (humorously?) wonder if any of you have automated AI using the git ~blame/history to email the humans who wrote some code asking for their rationales (or perhaps complaining?)

Comments URL: https://news.ycombinator.com/item?id=48155104

Points: 3

# Comments: 1

There were those used to try hard, carefully yet hopefully, to learn how robots, computers and "artificial intelligence" might benefit or harm people, especially for those with special needs.

But now I think we're in a harms race.

[FYI, Proto was a magazine published by Harvard Medical School's teaching hospital, Massachusetts General Hospital, from 2005-2022.]

Comments URL: https://news.ycombinator.com/item?id=47784843

Points: 1

# Comments: 1

  All that is gold does not glitter,
  Not all those who wander are lost;
  The old that is strong does not wither,
  Deep roots are not reached by the frost.

> You can easily extract a design system from any URL

[1] https://blog.google/innovation-and-ai/models-and-research/go...

In fact I took your topmost entry with no helpful site/update tags and dove in a little to try to understand why a RSS friendly blogger might not be passing along ~ tags for better reader discovery.

Turns out my scarce info test case blogger has a mastodon that immediately lists all these tags about himself [I've stripped it down] ...

#FrontEnd Developer #CSS #Halifax #London #Singapore Technical writer and rabbit-hole deep-diver Former Organiser for https://londonwebstandards.org & https://stateofthebrowser.com Interests: #Bushcraft #Outdoors #DnD #Fantasy #SciFi #HipHop #CSS #Eleventy #IndieWeb #OpenSource #OpenWeb

I conclude if he knew such site and post tags getting to RSS would be of use, he'd probably make the tiny effort to wire the descriptions.

Nonetheless I merely crawled links for a minute to found this info, so I imagine something like the free tier of the Cloudflare crawling api might suffice over time for a simplistic automated fix to hint decorate blog sites.

I mean, given that we're not trying to recreate pagerank, but just trying to tip the balance in favor of desirable initial discovery.

> March 15 there were 1,251 updates [from feed of small websites ...] too active, to publish all the updates on a single page, even for just one day. Well, I could publish them, but nobody has time to read them all.

if the reader accumulates a small set of whitelist keywords, perhaps selected via optionally generating a tag cloud ui, then that est. 1,251 likely drops to ~ single page (most days)

if you wish to serve that as noscript it would suffice to partition in/visible content eg by

  

  
  Prospero: Where in the world is my teapot? Hello? I'm waiting!