Scribblenauts is not LLM-based – its first game was released in 2009. The creators just spent a lot of time making a database of objects that people might ask for, their game properties, and what they look like in the game’s simple art style. Scribblenauts is also different in that it is a 2D puzzle platformer set in small levels that have their own goals, rather than a top-down MMORPG like Asciidia.

While I found the concept of Scribblenauts cool, I quickly grew bored of the puzzle levels and the shallow combat mechanics. It didn’t help that a Black Hole could solve almost any problem by destroying an obstacle. This game’s monetary cost for casting solves that, at least, though I don’t know if the economy would be a significant limitation in the long term.

Some people have griefed the starting area by placing turrets that kill your character when you’re in range. It makes for a confusing first 10 minutes when you wander around and aren’t sure why you’re dying. Even after that, it’s not clear to me how to interact with hostile structures, or how to tell which structure is on my team and will therefore attack me.

I tried creating a few weapons, but neither worked. I had a Short Bow equipped and Arrows in my inventory, but the Short Bow kept saying “out of ammo”. Same with a Torpedo Launcher when I had Photon Torpedoes in my inventory. When I created a plain Torpedo, I could shoot it, but it just flew off and disappeared without affecting the structure I aimed at.

There is an NPC asking for a boat, which is a cool idea for using the conjuring system, but I can’t get near him for long without dying to a griefer’s turret, so I don’t feel like trying to solve his quest.

Other hypothetical prediction markets whose insights would be useful include those used in futarchy, a proposed government system in which decisions are made based on betting markets. The proposal: https://mason.gmu.edu/~rhanson/futarchy.html; some analysis: https://www.lesswrong.com/w/futarchy. In futarchy, prediction markets would be set up for, for example, “average happiness of citizens (as measured by regular survey) will increase in 1 year if Bill ABC passes” and “average happiness of citizens will increase in 1 year if Bill ABC does not pass”. The government would pass or reject proposed bills according to whichever market predicts higher happiness, and the market describing the event that did not happen would be closed and its money refunded.

It’s the same here. For example, this study concluded that most changes are safe and some are very bad, as opposed to most changes being slightly bad. That is not obvious, especially to infrequent LLM users.

Also, even “obvious” conclusions are within the scope of science. I’ve spent too long writing this already to look up an example, but I bet there have been countries in the past whose leaders chose “obviously-good” monetary policies that economic research could have shown was counterproductive. The world is complicated, and without systems of communication such as academia, it’s hard to be sure if what you see is what everyone else sees.

  - Djot requires
  
    - writing nested lists
  
      - with blank lines in between
  
    - successive list items at the same level
    - can skip the blank line
  
      - but not this list item

Djot could have avoided the blank line requirement by not trying to join hard-wrapped lines back into one paragraph / list item. That would work for me because I only soft wrap my text. Djot’s choice to support hard wrapping caused all of its users (including those who hard wrap) to have worse nested list syntax.

  <<<<<<< left
  ||||||| base
  def calculate(x):
      a = x * 2
      b = a + 1
      return b
  =======
  def calculate(x):
      a = x * 2
      logger.debug(f"a={a}")
      b = a + 1
      return b
  >>>>>>> right

I also use the merge tool of JetBrains IDEs such as IntelliJ IDEA (https://www.jetbrains.com/help/idea/resolve-conflicts.html#r...) when working in those IDEs. It uses a three-pane view, not a four-pane view, but there is a menu that allows you to easily open a comparison between any two of the four versions of the file in a new window, so I find it similarly efficient.

Comments URL: https://news.ycombinator.com/item?id=47458284

Points: 4

# Comments: 0

• Which of axe-core’s rules (https://github.com/dequelabs/axe-core/blob/develop/doc/rule-...) LLMs violate most often

• Which groups of users are most affected by those rule violations (e.g. blind users or deaf users)

• Whether it’s likely that I unintentionally violate those same rules in web pages I write

Examples of rule violations and statistics on most-violated rules would make the website more convincing by showing that the detected accessibility errors reflect real problems. It would rule out that the only detected error was a single noisy false positive rule in axe-core. I bet that most readers are not familiar enough with axe-core to trust that it has no false positive rules.

Comments URL: https://news.ycombinator.com/item?id=46528502

Points: 26

# Comments: 7

A better comparison would use, for example, a React component:

  function Button({ children }) {
    return (
      
        {children}
      
    );
  }
  
  // Usage:
  Save

A better argument against Tailwind would be the added complexity of having a build system and a front-end framework or templating language, if your project doesn’t already have those for other reasons.

(adapted from my better-formatted comment at https://lobste.rs/c/oznzzj)

Also, from Jujutsu’s README (https://github.com/jj-vcs/jj#mandatory-google-disclaimer):

> I (Martin von Zweigbergk, martinvonz@google.com) started Jujutsu as a hobby project in late 2019, and it has evolved into my full-time project at Google, with several other Googlers (now) assisting development in various capacities.

Pronounced correctly, “segue” sounds just like “Segway” – not like “seg-oo”, as you might have assumed.