CatherineFlick

CatherineFlick, 1 month ago to random

I'm really hoping you don't need me to tell you that using machine learning ("AI") systems to identify potential targets and then to suggest that they be targeted when at home with their family members, including children, is one of the most abhorrent, unethical, inhumane things I've ever seen. There is absolutely no excuse for developing these systems. Technology is never neutral. https://www.theguardian.com/world/2024/apr/03/israel-gaza-ai-database-hamas-airstrikes

CatherineFlick, 10 months ago to random

I told everyone so. https://archive.is/EOt0J

CatherineFlick, 1 year ago to internet

So I’m watching over on #Bluesky they’ve “realised” they can train their own model for content moderation and are laughing because it’s flagged up “pestilence” as a word indicating a problematic post. Now I’m no ML whiz but surely they’re going to need a boatload of human work to get this going (tagging, testing, retraining) and it’s not just something to “mess around with” live as they seem to be doing?

Screenshot of Bluesky posts by Whyrusleeping@why.bsky.team who is training ML models. Alright, lets make some GPUs sing. Teaching some linear algebra how to do vibe checks Need more training data. Current model thinks that a post containing just the word "pestilence" has a bad vibe. Idk, maybe its right? pestilence okay cool, the linear algebra is getting better

CatherineFlick, 6 months ago to Electricvehicles

My 1 year old son has just started making noises when he plays with cars. He doesn’t say “brrrrm” or similar, he says “eeeee”. We own electric cars.

Think of it, a future where cars, buses, trucks, tractors go “eeeee” instead of “brrrrrm” in the minds of children.

#ElectricVehicles

CatherineFlick, 3 months ago to random

As someone who did her PhD on consent I just want to flag a few things in the bsky bridge discussion. One: medical style disclosure based consent (terms and conditions, EULAs etc.) is totally inappropriate for this sort of situation and has been for decades. But it serves the needs of slow-moving legal requirements and companies that like people to forget they signed up to stuff (or were coerced into doing so for social or other reasons). See https://liedra.net/thesis for more details. (1/n)

CatherineFlick, 11 months ago to random

I'm so confused by the boiler setup in this house (we didn't set it up). My visiting parents aren't getting hot showers in the evening, and I can't for the life of me work out why. To make matters more confusing we have an eddi which diverts solar energy to the immersion heater (but also indirect boiler heated?) tank. I can't work out how the boiler decides when to heat the cylinder (and whether it's the hot water or central heating system that does it).

CatherineFlick, 9 months ago to random

PSA to people still posting on Twitter: Twitter no longer shows replies or threads to non-logged-in users - just the individual tweets you link to. It also doesn't show profiles with all your recent tweets. I'm not sure this is widely known, especially for those in charge of brands etc. It really cuts down on the usefulness of Twitter IMO.

CatherineFlick, 1 year ago to random

Ok BBC iPlayer play from start, let’s go #Eurovision

CatherineFlick, 1 month ago to random

Can anyone explain to me how these people sleep at night? I just don’t understand how a company that feeds kids can see a profit rise of 30% as anything other than a damning indictment of the quality of their food and treatment of their workers.

CatherineFlick, 11 months ago to random

This statement on AI is buying into the longtermist/effective altruism philosophy that is really problematic. If AI poses an existential risk it's because of the people who make AI. "AI" is just a computer program. If Sam Altman et al really believe the risk they should, y'know, stop developing it, and, more importantly, stop profiting from it. Also, let's focus on the problems being caused by AI right now, the ones that a bunch of the signatories here have profited from (and will continue to).

CatherineFlick, 4 months ago to random

Just realised this is my last day as not a professor, tomorrow I reach that milestone with my new job!

CatherineFlick, 9 months ago to random

I mean normally I hate to say I told you so but in this instance no, no I don’t. https://arstechnica.com/tech-policy/2023/08/buyers-of-bored-ape-nfts-sue-after-digital-apes-turn-out-to-be-bad-investment/ (h/t @pikesley )

CatherineFlick, 10 months ago to random

Computer scientists know fuck all about school education. This one in particular assumes that all teachers do is regurgitate information. Why anyone would listen to this guy is beyond me. AI hype is so ridiculous https://www.theguardian.com/technology/2023/jul/07/ai-likely-to-spell-end-of-traditional-school-classroom-leading-expert-says

CatherineFlick, 1 month ago to random

I had to look up what the Papa Johns pizzas bought for 10,000 bitcoin were "worth" today. oof.

CatherineFlick, 7 months ago to random

Same, buddy, same.

CatherineFlick, 11 months ago (edited 11 months ago) to random

Harvested some garlic today as the leaves were getting pretty dead. Looking good! Just reading loads of conflicting info on how to dry it. Greenhouse? Outside? Heat? Cool? Sun? Shade? Plait? Lie out solo? #garlic #allotment I tried an experimental one last week in the greenhouse and it went yellow and soft? And there doesn’t seem to be much in the way of cloves.
Edit to add: this is autumn planted garlic

CatherineFlick, 8 months ago to random

WTF does this even mean??? “Downing Street is spending £100m on a new AI taskforce to help test algorithms as they are developed. “
That’s…. not at all useful??? https://www.theguardian.com/technology/2023/sep/25/ai-bioweapons-rishi-sunak-safety

CatherineFlick, 1 month ago to random

So Kagi has pivoted to suckage, DDG is kinda rubbish now, and lol Google, what are the cool kids using to actually find stuff on the internet? Are we back to blogrolls and web rings?

CatherineFlick, 5 months ago to random

Turns out our solar together (booooo) engineer was telling us a bunch of porky pies when he told me that we could run the house off the battery in the case of a power outage! Grrr!

CatherineFlick, 1 month ago to random

I was critical of the Cass Report earlier, talking about how discounting studies that weren't double blinded RCTs was bad form. Apparently Dr Cass has responded, saying that '"obviously" young people could not be blinded as to whether or not they were on puberty blockers or hormones because "it rapidly becomes obvious to them [...] But that of itself is not an issue because there are many other areas where that would apply"' https://www.bbc.co.uk/news/health-68863594

CatherineFlick, 8 months ago to random

Very excited to announce I’ll be moving to Staffordshire University in the new year as a Professor of Ethics and Games Technology!

CatherineFlick, 10 months ago to random

Just had a massive rant to our original solar installers about how bad the Solar Together people were. Still angry about the whole thing where they replaced our perfectly good inverter with one that isn’t as good because their batteries didn’t work with it (apparently they could have used different batteries that did work with the original setup). Anyway avoid the #SolarTogether program if you’re in Leicester. The company that does it is bad bad bad.

CatherineFlick, 3 months ago to random

Apparently it now constitutes bullying and harassment to call transphobes transphobes in academic workplaces in the UK. I guess all the racists, misogynists and homophobes can come out and claim that their "genuine beliefs" should be respected too. Ugh.

CatherineFlick, 3 months ago to random

"Sora serves as a foundation for models that can understand and simulate the real world, a capability we believe will be an important milestone for achieving AGI." I can't believe they're still claiming that generative models can "understand" the world. They can't understand anything. They are just a fancy statistical model for correctness probability against previous known good outputs. You don't get AGI from that. You can't get AGI from that. Irresponsible claims.

https://openai.com/sora

CatherineFlick, 1 month ago to random

Why, for the love of the gods, is there a separate Microsoft Teams for "work or school", I mean this is just sloppy software design. Also, lol at "Microsoft Teams is generally available" on the download site. I mean I'd like it to be always available but if that's the best you've got...