![Toby Murray Profile](https://pbs.twimg.com/profile_images/1630752435592638467/CEd8gEyu_x96.jpg)
Toby Murray
@tobycmurray
Followers
1K
Following
8K
Statuses
5K
Go find me at @tobycmurray.bsky.social. Assoc Prof at @UniMelb and @cis_unimelb cyber lead; @DSI_VIC Director; @UniofOxford DPhil (@HertfordCollege); He/him
Melbourne
Joined February 2013
@AlexP_DownUnder They seem to be making design choices that may lead to better (less toxic) experiences for people
1
0
1
A cautionary tale also with relevance to those writing LLM powered “agents”
Letting chatbots run robots ends as badly as you'd expect via @theregister
0
0
1
@jer_gib @AndrewCMyers Especially not among graduate students. Or at many of the more progressive colleges.
1
0
2
If you want to understand why “predict when people will do crimes” is utter snake oil, check out @random_walker and @sayashk’s excellent book “AI Snake Oil”. Along the way you’ll learn also why not all AI is snake oil and how to sort the wheat from the chaff. Highly recommended
They already have our medical records, now Palantir ( run by Peter Thiel) want to get into the predict-when-people-might-commit-a-crime business. Thiel’s protege is JD Vance. Apart from that, have a lovely Saturday.
0
1
5
@adambspencer A mathematician slandered by a cryptography-related error message. What is the universe trying to tell us?
0
0
1
Minimum age of criminal responsibility in Australia: 10 years ACT: 12 NSW: 10 NT: 12 (for now) QLD: 10 SA: 10 TAS: 10 VICE: 10 WA: 10 Proposed minimum age to legally use social media: 16
No SCREEN TIME if you’re under 16 but you can do REAL TIME if you’re over the age of 10. Get real @AlboMP.
0
2
5
RT @AndrewCMyers: Discovery is the reward. No one understands anything fully. Optimism pays off. Everyone has skills you lack. There are no…
0
16
0
A great and timely initiative. Props to all involved
The review process is under threat and we must make it resilient against adversarial attacks. Proud to be part of @TheOfficialACM PROTECT where we'll look into securing academic reviewing
0
0
2
RT @tmiller_uq: I'm hiring for a postdoc position (women-only applicants) to work with the Ida Asadi Someh and Tapani Rinta-Kahila on desig…
0
6
0
To paraphrase Einstein, let me pose a provocation: As far as mathematical specifications refer to real software they are not certain, and as far as they are certain they do not refer to real software.
I absolutely love the pdf example. It tells us that in the “real world” specs don’t really exist and we shouldn’t expect them to. If you are able to precisely specify correctness for your software, it means your software is weird (highly atypical, not pejorative).
1
0
2