Here's my growing list of writing guidelines in my prompts. I feel that the RLHF isn't working well on writing? I have to fight slop aggressively. Rule of thumb: delete 50% of what AI writes because at least that much is low value fluff (but AI is still helpful) 1. Do not

My lab Johns Hopkins University is recruiting research and communications professionals, and AI postdocs to advance our work ensuring that AI is safe and aligned to human well-being worldwide: We're hiring an AI Policy Researcher to conduct in-depth research into the technical and policy

This paper shows the same effect as other studies of "cheating" with AI - if you use AI to do the work (as opposed to using it as a tutor), you don't learn as much. But note: the results are specific to the essay task - not a generalized statement about LLMs making people dumb.

Wrapped up Stanford CS336 (Language Models from Scratch), taught with an amazing team Tatsunori Hashimoto Marcel Rød Neil Band Rohith Kuditipudi. Researchers are becoming detached from the technical details of how LMs work. In CS336, we try to fix that by having students build everything:

I really like the term “context engineering” over prompt engineering. It describes the core skill better: the art of providing all the context for the task to be plausibly solvable by the LLM.

This study is being massively misinterpreted. College students who wrote an essay with LLM help engaged less with the essay & thus were less engaged when (a total of 9 people) were asked to do similar work weeks later. LLMs do not rot your brain. Being lazy & not learning does.

If you haven't yet, take a few minutes to fill out our SRL survey—your responses help us fulfill our mission of understanding and addressing the plight of SRLs in Canada! Find it at this link: surveymonkey.com/r/2016Intake

The Ontario Superior Court has a snazzy new website! *Searches around for court data* Oh. ontariocourts.ca/scj/about-the-…

“Since the judicial system exists only for them, and because of them, the citizens of this country must know and understand that towards which they pay so much money. After all…the public has a right to expect a fair return on its investment”. Justice L’Heureux-Dubé, Alta L Rev

The Kawaskimhon Moot is a national, multi-party moot that simulates negotiations focused on Indigenous legal issues and peoples in Canada. Learn more about this year's moot in a new blog post by Allard Law Professor Scott Franks. allard.ubc.ca/about-us/blog/…

I asked my wife why she was in private browsing mode on her phone and she explained that Safari only allows 500 tabs in regular mode so she had to switch. You think you know a person

Worth a read. Provocative fiction can do a lot to unstick conversations that have gotten slowed down by very legitimate multifarious nuances.

Completely horrifying.👇 Dark Age 2.0 "Only 35% to 40% of US pregnant women & parents of young children say they intend to fully vaccinate their child..." Most US pregnant women, parents of young kids don't plan to accept all recommended kids' vaccines cidrap.umn.edu/childhood-vacc…

An increasing number of Zoom bombings in Ontario courtrooms now means the public is banned from attending proceedings virtually. The assistant deputy attorney general said that "if the public would like to attend, they will be required to do so in person": cbc.ca/news/canada/wi…

Love this.

ICYMI The Ontario Court of Justice released a new policy that would stop observers from accessing court proceedings online unless they receive authorization from the judge overseeing the case. Those interested are encouraged to show up in person: globalnews.ca/news/11304457/…

I am a female criminal defence lawyer. I have 4 degrees. I articled at a leading Canadian defence firm and clerked at the Ontario Court of Appeal & the Supreme Court of Canada. I am not a skirt to hide behind. I earned my place in the courtroom. criminallawyers.ca/why-women-lawy…

I'm doing a trial today with a wonderful judge, a wonderful crown, and a wonderful client. It's a wonderful way to end the week. Be nice to each other. It's the best way to get through life.

The state-of-the-art in prompting is still much more art than science. Very few people are rigorously testing prompt approaches and almost everything you are taught about prompting is best guesses or based on obsolete information (eg chain of thought no longer helps much)

Two provincial court judges in Ontario stepped down last year before investigations could be completed into allegations of inappropriate and unprofessional behaviour, according to an annual report of the Ontario Judicial Council: thestar.com/news/ontario-j…