Peer review is for humans. Because my peers are human.
What's being proposed is robot review. It is fundamentally different than peer review.
Part of the goal here is to stem the tide of AI-slop submissions... so basically robots reviewing robots.
It's basically an arms race between LLMs.
'Announcing the move...Springer Nature said the project would explore how AI could help create a more “rewarding reviewer experience”, streamlining analysis of large volumes of text and figures, freeing up researchers’ time to allow them to focus their expertise where it is most needed.'
The solution isn't to put more LLMs in the mix.
We need to finally admit that the publication model for academic success is flawed and has been flawed for decades. LLMs just made it impossible to ignore.
What we have right now is a hot dog eating contest where the prize is more hot dog eating.
The LLMs will extrude more hot dogs and will grind them up to make more hot dogs.
In the end, all you will have are tasteless mush.
Academics need to find a new approach to work and reward -- one that is resilient to AI gaming.
The DARPA roots go back a long, long way.
Anyone who wasn’t comfortable knowing that the military would come calling should have done their homework.
It’s a little like SpaceX employees or Tesla owners suddenly protesting Elon when they discover what he’s really about.
“The city release of documents confirmed a police drone flew through the windshield of a Peterbilt tractor-trailer truck in June, cutting the face of a Canadian driver and sending him to Baylor Scott & White Medical Center-Temple.”
This is a key reason why I’m opposed to these drone 911 programs.
Dear university management,
You told us that we had to focus on in-person teaching after the COVID lock-downs. How about we continue to focus on people rather than AI-everything?
Seems like at least company is banking on a "more human" approach. Seem likes a good idea.
I ordered the 5Pro because I wanted an efficient whole room air purifier that is packable. I've already got a number of air purifiers I can fit in a suitcase, but none of them have really high airflows like the 5Pro does with it's 4 filter, CR Box-style configuration.
youtube.com/shorts/_rcbp...
Probably the coolest thing about being a prof that I didn’t expect when I started in 2008 is having the opportunity to learn from incredibly talented and smart students.
The sharing of knowledge as a social interaction is an incredible thing.
And whenever AI gets in the middle of that I feel sad for the lost social opportunity.
Went to get my car's oil changed and the service guy tried to upsell me by pointing to "recommendation" by Google's Gemini.
Not impressed.
I care what a qualified expert has to say. I care what the user manual for the car says.
I don't care what the stochastic parrot has to say.
My wife took a First Aid course a few days ago.
Her classmate-partner in the course was asked what he would do when trying to control a bleeding wound.
His answer? "Uh, first I'd check with ChatGPT"
Really? In an emergency? After taking the course?
Cognitive offloading at its "best".
Opinion: If AI is now part of academic writing, what should institutions actually evaluate? This essay argues that analytical writing should be judged by evidence, structure, argument, and accountability not by guesses about how each sentence was produced. https://chroni.cl/4eDRdjg
Opinion: If AI is now part of academic writing, what should institutions actually evaluate? This essay argues that analytical writing should be judged by evidence, structure, argument, and accountability not by guesses about how each sentence was produced. https://chroni.cl/4eDRdjg
Brad Bradford is on CBC telling the host that his construction plans require taking on debt and making our children pay later. Why? Because "it's not fair" to make today's residents pay for stuff that they'll only get to use later.
Pass the buck. On construction, the environment, the economy...
The rotting corpse of Canada's national magazine saw fit to print a first-person account of how a 23-year-old Alberta separatist was radicalized because a mask fogged up her glasses during COVID.
Honest to Zeus, this is the dumbest fucking timeline...
Maybe universities shouldn‘t be tying which departments and programs they decide to invest in based on projected job trends they have no actual clue about. Why not just invest proportionally in every field of inquiry rather than underfund some disciplines and overfund others?
Local woman wakes up, thanks 76% of Alberta for not being fucko buckos but reminds us that as we get the results from the rest of the 9 questions, this province still has a long way to go developing critical thinking skills and unlearning systemic racism, makes oatmeal.
Canada's province of Alberta has rejected a plan to hold a binding independence referendum, results show, a setback for separatists backed by allies of US President Donald Trump
u.afp.com/SCah
Canada's province of Alberta has rejected a plan to hold a binding independence referendum, results show, a setback for separatists backed by allies of US President Donald Trump
u.afp.com/SCah
We also need non-stagnant funding for discovery research. And we need leaders like Carney (and Harper before him) to stop directing huge amounts of money to recruiting a handful of "stars" and supporting their favourite industries like fossil fuels and AI.
Canada has considerable strengths across its research ecosystem, but lacks the coordinated planning needed to fully realize its value. go.cca-reports.ca/4hWNe3z
On CBC they’re reporting how “[Meta’s] Muse doesn’t check with me… and gives some guy my address.”
Expect more bad stories about people using AI like this.
blogflow.ca/en/article/y...
We need whole cross discipline, college level, courses on how the human mind is primed for this.
How we are forged to form community, how we seek a deeper meaning to our existence, how we Want To Believe so fucking much.
We were built to fall for this.
We are in so much trouble.
Honestly, this is worse than NFTs and cryptocurrencies.
People I really thought would know better, are falling for this.
Your brain is fucking with you.
You’re seeing Jesus on toast.
LLMs produce statistically plausible text given a particular context.
«Agents» run that in a loop, «given everything we’ve said so far and this new text, what is the most plausible next piece of text?»
There is nothing there.
It’s all in people’s heads.
The future is in reach. More of this, and a focus on perfecting battery technology as we transition from AC dominance to DC dominance. Such a shift would change EVERYTHING.
The Canadian military had better not be designing systems to rely on Musk’s Starlink / Starshield.
The Americans applied the kill switch to it… degrading Ukrainian defence systems in the process.
Doesn’t matter who applied the switch. The fact is they did.
#ElbowsUpwww.npr.org/2026/10/18/n...
LLMs produce statistically plausible text given a particular context.
«Agents» run that in a loop, «given everything we’ve said so far and this new text, what is the most plausible next piece of text?»
There is nothing there.
It’s all in people’s heads.
After over two years living in Montreal, I can safely say the family doctor situation is absolutely fucked.
I FINALLY got a letter that I got one.
But after weeks of trying, I haven't been able to contact them. Phone? Hangs up, no voicemail. Email? No reply. In person while 'open'? Door locked.