Rich Lafferty

Rich Lafferty

@rich.lafferty.ca

Disaster prevention chihuahua. Reliability at PagerDuty. Living on unceded land in Punamue'katikt, Mi'kma'ki (Dartmouth, NS). May all beings be free from suffering. He/him. also hachyderm.io/@mendel. formerly @mendel on twitter. www.lafferty.ca

1.1K followers885 following4K posts

Pinned

Rich Lafferty
@rich.lafferty.ca

Well this thing seems to be getting popular. Intro post(s) time! Professionally, I'm a staff reliability engineer (#SRE) at PagerDuty, where my interests lie in #LearningFromIncidents, #HumanFactors and safety. Personally... (1/2 - cat pictures to follow)

Rich Lafferty
@rich.lafferty.ca

quite possibly the last good motorcycle day of the year BUT i just got my flu and covid shots and while i’m not sore yet i know my body is gonna be angy soon (even with prophylactic desloratadine)

Rich Lafferty
@rich.lafferty.ca

yeah body angy now. fatigue, chills the problem with the antihistimane thing is that it only helps with covid vaccine reaction, not flu vaccine

Rich Lafferty
@rich.lafferty.ca

CLAUDIFORM

Erin Kissane@kissane.myatproto.social

One by one, the independent artists whose shop newsletters I subscribe to have switched to Claudiform emails that all sound the same: The vase by the bedside. The quiet moment before sleep. A tea blend that reminds us to pause. It’s such a bummer.

Rich Lafferty
@rich.lafferty.ca

quite possibly the last good motorcycle day of the year BUT i just got my flu and covid shots and while i’m not sore yet i know my body is gonna be angy soon (even with prophylactic desloratadine)

Rich Lafferty reposted

rahaeli
@rahaeli.bsky.social

This is a fascinating paper about neurolinguistics

Andrew Ducker @andrewducker.bsky.social

Why is AI-generated language so hard to understand?

arxiv.orgWhy is AI-generated language so hard to understand?There is a widespread sense that AI-generated text has recently become harder to understand. In this study, we investigate a set of psycholinguistic features that are known to correlate with human processing difficulty (e.g., object relative clauses, long syntactic dependencies, low-frequency n-grams). Using a novel corpus of AI-generated text across diverse domains, we find that AI models use these difficult features more often than paraphrases generated by round-trip translation. Compared with human writing, all nine frontier and near-frontier models use more inanimate subjects with animate verbs and more low-frequency n-grams, and most also use longer dependencies, more stacked nouns, and more reduced object relatives in conversation; we also demonstrate substantial variation across model families. We propose and test a hypothesis as to why: while human language is optimized for successful communication, AI-generated text is increasingly optimized for other goals like math, code, and reasoning, in which there is a pressure for \textit{information density} rather than successful communication. We provide a proof-of-concept for this hypothesis by fine-tuning a language model under a length penalty, which leads to a divergence from human-like linguistic behavior. Our findings highlight the axes of misalignment between linguistic processing demands in humans and LLMs, and point toward possible remedies.
Rich Lafferty
@rich.lafferty.ca

A thing LLMs can do pretty well is blazon heraldic devices from a picture, and to emblazon an image of the arms from its blazon. that’s neat in an anachronistic kind of way.

Rich Lafferty reposted

Molly White
@molly.wiki

truly a wild time to be alive when the cryptography world is split between those worried about quantum computers and those whose private key is “4”

Rich Lafferty
@rich.lafferty.ca

in queue for sailgp halifax tickets!!! annoyingly the grandstand is split into three price levels this year. can't see prices yet until they open the doors at 10 (no george's island this year either, but not surprised, that didn't seem like a great viewing spot AND you're stuck there)

Rich Lafferty
@rich.lafferty.ca

Ah but it does not require a new tzdb update — NO has used America/Winnipeg all along, and it would’ve required an update if they *didn’t* follow Manitoba!

Rich Lafferty
@rich.lafferty.ca

oop, a fuse on the power pole* across the street from us just let out the magic smoke! we lost a leg, main breaker off now, hopefully it didn’t fry anything. (* almost said “hydro pole”, tough habit to break!)

Rich Lafferty
@rich.lafferty.ca

and we’re back! power adapter on our network media converter (HH3k bypass) was unhappy but luckily i had an equivalent in The Parts Bin good thing i held onto all those old cables and chargers!!!

Rich Lafferty
@rich.lafferty.ca

am I going to regret an all-day 12ºC motorcycle ride tomorrow, when it might be my last chance at a long ride for the season? LET’S FIND OUT also, one of these days i need to get heated grips

Rich Lafferty reposted

rahaeli
@rahaeli.bsky.social

Replying to @rahaeli.bsky.social

bsky.app/profile/raha... is the specific pre-vaccination prophylaxis post in that thread if you need it and don't feel like wading through all the rest, heh

rahaeli@rahaeli.bsky.social

The dosage and timing for pre-vaccination is EITHER: - 50mg diphenhydramine (Benadryl) OR 50mg promethazine (Phenergan) 15 minutes before the shot OR - 20mg cetirizine OR 20mg loratidine 2-3 hours before the shot Only take one of those (ie, don't take Benadryl AND cetirizine).

Rich Lafferty
@rich.lafferty.ca

i have a new concrete pad in front of our garage door.  the concrete is hard, thankfully, but it still has the wood frame around it because it was only poured yesterday.  AMAZON LEFT A PACKAGE ON IT