Login

Recent searches

No recent searches

Search options

Not available on writing.exchange.
writing.exchange is one of the many independent Mastodon servers you can use to participate in the fediverse.
A small, intentional community for poets, authors, and every kind of writer.

Administered by:

Server stats:

323
active users

writing.exchange: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#errors

1 post1 participant0 posts today
Public
seanpm2001 🇺🇦️

If there are any OpenStreetMap contributors reading, I have an important correction that needs to be made regarding Spokane, Washington:

I used to live on the South Hill, and could walk to Moran Prairie Elementary School within 10 minutes. It was located at the base of the hill (the street I used to live on is highlighted in orange) However, on OpenStreetMap, it is showing Chase Middle School, which is much further away. I don't know how to edit OpenStreetMap, but there is an important correction that needs to be made here.

I have noted this error to be visible for over 2 months without any attempts to fix it.

#openstreetmap#osm#spokane
Continued thread
Public
Zhi Zhu 🕸️

DOGE’s credibility is shot.
robertreich.substack.com/p/thi

"DOGE has saved the govt just $2 billion. That comes to 1/35 of 1% of the federal budget.

Repeated #errors have raised questions about the quality & veracity of the #information Musk’s #DOGE is putting out. The mistakes also call into question the team members’ competence — whether they understand the #government well enough to cut it while avoiding catastrophe."

Sub-heading and text from article (edited for length): 3. DOGE’s credibility is shot. ... Not even these contract cancellations amount to much. The New York Times found many so-called “cancelled” contracts had actually ended under previous presidents. By some estimates, DOGE has saved the government just $2 billion. That comes to 1/35 of 1 percent of the federal budget. Repeated errors have raised questions about the quality and veracity of the information Musk’s DOGE is putting out. The mistakes also call into question the team members’ competence — whether they understand the government well enough to cut it while avoiding catastrophe. This week, Musk again demanded every federal worker justify their employment, and once again, his hastily executed email has sowed confusion and resistance throughout the workforce, with many agency heads openly defying it. Federal workers at NASA’s Human Landing System (HLS) program didn’t know how to respond because HLS has contracted with Jeff Bezos’s Blue Origin rocket company, the major competitor to Musk’s SpaceX, and HLS workers didn’t want to disclose to Musk what they were doing. At the same time, the Commerce Department has adjusted a government program to make it easier for states and municipalities to use government subsidies to buy Musk’s Starlink internet service. (Mark my words: When all this is put together, it will make Warren G. Harding’s Teapot Dome scandal look like a game of checkers.)
#ElonMusk#Musk#Coup
Public
Nick Radcliffe

Bruce Schneier perfectly articulates something I’ve been thinking about:

AI Mistakes Are Very Different from Human Mistakes
schneier.com/blog/archives/202

This is exactly right, certainly for LLMs.

"It seems ridiculous when chatbots tell you to eat rocks or add glue to pizza. But it’s not the frequency or severity of AI systems’ mistakes that differentiates them from human mistakes. It’s their weirdness. AI systems do not make mistakes in the same ways that humans do.”

Required reading.

Schneier on Security · AI Mistakes Are Very Different from Human Mistakes - Schneier on SecurityHumans make mistakes all the time. All of us do, every day, in tasks both new and routine. Some of our mistakes are minor and some are catastrophic. Mistakes can break trust with our friends, lose the confidence of our bosses, and sometimes be the difference between life and death. Over the millennia, we have created security systems to deal with the sorts of mistakes humans commonly make. These days, casinos rotate their dealers regularly, because they make mistakes if they do the same task for too long. Hospital personnel write on limbs before surgery so that doctors operate on the correct body part, and they count surgical instruments to make sure none were left inside the body. From copyediting to double-entry bookkeeping to appellate courts, we humans have gotten really good at correcting human mistakes...
#AI#LLM#Errors
Public
gprimiero

A while a go I started working on a modal semantics for defining errors in distributed computing via types. The results have now appeared in a chapter for a volume dedicated to one my PhD mentors and teacher Göran Sundholm. The actual formalism is a bit far away from the tons of things I learned from him, but it touches on the issues of correctness and errors in proofs part of his philosophical research.

#proofs #computing #errors #correctness

link.springer.com/chapter/10.1

@philosophy

SpringerLinkHandling Mobility Failures by Modal TypesCorrectness is a major concern for logical systems, especially for its significance in computational settings. While establishing a norm for the correctness of computational procedures is a standard requirement, defining errors is a less investigated formal problem....
Replied in thread
Public
Daniele de Rigo

5/

Subtle tradeoff "whether avoidance increases for more difficult instances, as would be appropriate for the corresponding lower level of correctness"

Alas, the
"percentage of avoidant answers rarely rises quicker than the percentage of incorrect ones":
"an involution in #reliability: there is no difficulty range for which #errors are improbable, either because the questions are so easy that the model never fails or because they are so difficult that the model always avoids giving an answer"

Public
Melroy van den Berg

@laurenshof Hoi Laurens. We zagen een vervelend probleem met parsen van de JSON van jouw blog AP nodeinfo API: browser.pub/https://fediverser

Zoals je ziet in de JSON output returneert deze API activeMonth & activeHalfyear als een string ("2") in plaats van een integer (2). Welke Wordpress plugin gebruik je? Dan meld ik dit probleem upstream.

browser.pubhttps://infosec.exchange/users/USBTypeSteve · BrowserPub · A browser for exploring ActivityPub and the fediverseExplore the open social web through the lens of ActivityPub and the fediverse.
#bugs#activitypub#parse
Replied in thread
Public
Kent Pitman

@screwtape

Well, the totality can be a function of other than args, so something that takes any kind of args can still err for other reasons. I don't know if that helps.

I don't know if you've read these papers but you might find them a bit more abstract:

nhplace.com/kent/Papers/Except

nhplace.com/kent/Papers/Condit

The second of those is mostly more advanced topics, as I recall. But the first is a discussion of the same issues as you find in Revision-18.txt, but in a way that is hopefully more abstract. That 1990 paper won a Best Paper award at the First European Conference on the Practical Application of Lisp, so I assume it was decently readable. It was just a dusting off of the paper I'd written in 1985 that no one had really cared about for 5 years. Since the paper was already written, I used the prep time for the conference to polish, and to whittle away anything about the paper that might not look like a "best paper", which apparently worked. :)

The 1985 paper is here, though I recommend just reading the 1990 version. I just often speak in hypertext and feel compelled to link things I talk about:

dspace.mit.edu/bitstream/handl

#lisp #CommonLisp #exceptions #errors #ErrorSystem #ConditionSystem

cc @masso

ExploreLive feeds
About