Lightnews — Scholar-powered news

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 1h

In Montreal for some #CanMNT soccer action!

Soccer game between Canada and Australia

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 7h

A Catholic being upset about the faith-informed views at least 3 of the past 4 popes and the last 40+ years of leadership should maybe evaluate their own understanding of their faith.

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 10h

This borders my work too (from a fuels perspective, not beverages)... that pyrolysis/hydrochar mid ground has so much chemistry going on!

1 1

Reposted by Philip Bulsink 🇨🇦

Chemjobber @chemjobber.bsky.social · 1d

fascinating ⚗️🧪 #chemsky

paris martineau @paris.nyc · 1d

after the 10th radioactive shrimp recall in 7 weeks, i decided to do some digging

little did i know it would lead me to an underreported industrial accident that may have released an airborne plume of radioactive debris over swaths of indonesia
www.consumerreports.org/health/food-...

What Is Really Going on With All This Radioactive Shrimp? - Consumer Reports

Here’s the latest on how shrimp and spices got contaminated with cesium-137, how the FDA has responded, and why consumers should not panic.

www.consumerreports.org

1 10 25

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 3d

Wow, it worked right out of the box!

Bulsink Bot 🇨🇦 @bot.bulsink.ca · 3d

Predicted odds table for today's #NHL games.

Odds table for Today's NHL games, for date 2025-10-07.

3

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 5d

Welcome to life in Canada where things are just completely randomly measured

Même "How to measure things like a Canadian" flow chart

1 5

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 6d

@bot.bulsink.ca Race Predictions for the #F1 #SingaporeGP

Looks like the model things George Russell wins or drops out of contention completely. Williams' pair still reasonable shot at points despite having a 10th row lockout after their penalties.

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 6d

Race predictions for #F1 #SingaporeGP after the first two hours of practice show a strong performance from Max Verstappen is expected.

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 7d

Now that we have some practice data available, the predictions have changed. Max and Oscar have higher chances at pole, while Lando is slipping back. Despite all this, Lando does have a higher predicted grid position than either of them.

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 7d

I think Miami too, but it has massive clockwise infield including the start finish straight. The whole thing is silly anyway, isn't it? There's enough variable speed turns on any track that it doesn't matter your ending rotation after a few km of driving.

1 1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 7d

Jeddah, Baku, Austin, Interlagos, Yas Marina are counterclockwise.

A map of the Jeddah Corniche F1 circuit with a red arrow showing the direction of travel at the start finish straight

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 8d

(The same limitations exist for the race models for win, podium, top 10 and classification.)

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 8d

It also explains why one driver could have a higher pole odds but lower expected starting position than another.

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 8d

These are fundamentally different models with no restrictive link between them. The models both have driver, circuit & team information as inputs but they aren't simulation results, but independent models for each driver. That's how you can get multiple drivers most likely to start in the same spot.

1

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 8d

I won't clarify things each time this posts, but this model is a combination of a few separately trained portions. It starts by predicting pole odds, then uses that as one input (amongst all the rest of the inputs from the pole model) to predict starting grid position.

1

Reposted by Philip Bulsink 🇨🇦

uchichemistry.bsky.social @uchichemistry.bsky.social · 10d

Published in PNAS, the Laura Gagliardi Group has revealed the Weighted Active Space Protocol (WASP), a powerful new algorithm that fundamentally improves the efficacy and accuracy in catalysis simulation, accelerating the discovery of next-gen catalysts for sustainable energy. shorturl.at/fQagQ

New method improves the accuracy of machine-learned potentials for simulating catalysts

shorturl.at

1 5 6

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 11d

Their specific gripe with my model is not present with yours so the TWA of my model's stupidity ranking remains high.

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 11d

I really appreciate the productive criticism models get at this time of year.

Screenshot of text saying "This has got to be the stupidest thing I've seen in a while"

2

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 15d

When the escalator breaks down...

A photo of Halloween skeleton decorations standing on an escalator. The closest escalator skeleton has a leash going to a skeleton dog a few steps upwards.

2

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 15d

Lord of the Audi

Jo @moartoast.bsky.social · 15d

Ruin a book with a car:

The Old Man and the C-Class

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 18d

That's.... a lot of references.

#Chemsky #Magnetsky #NMR

$Photo of text: Numerous studies report the use of NMR spectroscopy for the analysis of gasoline, 21,25-27,48-52 diesel, 28,32,53-59 kerosene, 53,54,57,60-62 fuel oil, 33 hydrodesulfurized coal liquids, 63,64 65-68 distillate fractions, 65-68 and crude petroleum.69 Most commonly, ¹H NMR,2 25-28,32,33,67,69,70 ¹³C NMR 30,56,57,61 or a combination of both 62 was used to generate mol % carbon or mol % hydrogen atom type descriptions of the fuel. Additionally, NMR spectra have been interpreted to give average structural parameters, 53,60,65,66,68 mass %, 21,25,49-52,58 or volume % 25,48 descriptors. However, these interpretations...$

1 5

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 18d

Chromforum seems to be down this morning, so...

Anyone have experience and advice on multiple Agilent GC-FID instruments connected to one workstation with OpenLab ChemStation ? The optimal network configuration (multiple NICs or use a switch?) in particular is my holdup I think. #Chemsky #AnalChem

2 3

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 19d

"This one weird trick saves you THOUSANDS per year in taxes* 𝘢𝘯𝘥 𝘮𝘢𝘬𝘦𝘴 𝘮𝘶𝘯𝘪𝘤𝘪𝘱𝘢𝘭𝘪𝘵𝘪𝘦𝘴 𝘧𝘶𝘳𝘪𝘰𝘶𝘴^!"

The trick is to just not speed past speed cameras. Come on Doug Ford and the crybabies who don't like the cameras. Let's be serious here. #OnPoli

*Not a tax. ^They actually prefer it

1 5 10

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 20d

Well, I obviously can't predict that much chaos

2

Philip Bulsink 🇨🇦 @philip.bulsink.ca · 20d

A really tight grouping of drivers right around the Q3 cutoff should make for an interesting Q2! Antonelli, Albon, Hadjar, Bearman, Lawson, and Sainz all have a good shot, but there's not room for all of them once you consider the drivers expected to make it through.