OhNoMoreLemmy
@OhNoMoreLemmy@lemmy.ml
- Comment on Blacksky Is Nothing Like Black Twitter—and It Doesn’t Need to Be 1 week ago:
I mean the fundamental problem is that humans are dicks and moderation is always needed. It should also be paid, and supported with counciling and recovery time when needed. Dealing with toxic content is a job.
Federation isn’t very good at this. The tech is great but everyone is a volunteer and there’s (afaik) no global ban hammer so trolls move from one instance to another. Bluesky currently has venture capital to pay for moderation teams, and centralized ban options.
I don’t know how long this can last without advertising revenue though.
- Comment on Whatsapp just developed a new murder weapon 1 week ago:
The other great ambiguity is hug/jazz hands. 🤗
I’m sorry your mom just died. *jazz hands*
- Comment on Is there a way to promote a community you create without stepping on toes? 3 weeks ago:
Ok, it’s almost empty at the moment.
The best thing you could do is post articles to it at a rate of about 1 a day. This will turn up in people’s feeds and hopefully get up voted and attract comments.
Cross posting etc., can come after this. You can’t promote a community with no content.
- Comment on Neurosurgery on Saturn 4 weeks ago:
- Comment on NHS-branded baby formula could prevent parents paying too much, watchdog says 1 month ago:
Just calm down and wait. You don’t actually need massive amounts of milk from start, and it takes time for everything to kick in.
- Comment on NHS-branded baby formula could prevent parents paying too much, watchdog says 1 month ago:
The important thing about NHS care is that half the midwives are completely insane, and they all contradict each other. They have basically no medical training and are just meant to get someone qualified if anything goes wrong. Instead they go mad with power and use it to bully first time parents into doing what they say.
Our midwife was insistent that if there wasn’t enough breast milk immediately after birth we had to switch to bottlefed. I don’t think you can figure out the official NHS approach based on anything a midwife says.
- Comment on What Sank the Bayesian Superyacht in Italy? 1 month ago:
I guess you could say puts on sunglasses it’s posterior collapsed.
- Comment on A decline in arable land 2 months ago:
The infographic says crop farming. Ireland is green because of the grass which goes with animal farming.
The chart is lazy nonsense that ignores most of the farming in Ireland.
- Comment on Math Research 2 months ago:
Fucking nailed it. My favorite application of them is to brew espresso at exactly the right temperature.
- Comment on What letter has the best games? 2 months ago:
And all the doom games.
- Comment on Anon quits their job 2 months ago:
In theory. It’s just standard contract law. You violate the contract, so you have to make the other party right.
In practice, the court is likely to go, “You should’ve hired someone else to do the work. No costs”
- Comment on Anon quits their job 2 months ago:
No, not at all.
If the company fire you they have to pay you, e.g., three months notice, regardless of if they want you to do the work or not.
If you quit without notice, you might have to pay the costs incurred by you quitting early, but that’s not your salary -because they wouldn’t be paying you in the first place.
Costs might be something like having to refuse an order because you now don’t have enough people to do the work, or the increased cost of an expedited hiring process.
I don’t know how common costs are in France, but the UK has the same rules and essentially no one ever claims costs. You need to really fuck over your employee in a very explicit and well documented way for this to even be considered.
The main disadvantage is you will have a bad reference if you leave without notice.
- Comment on Can AI talk us out of conspiracy theory rabbit holes? 3 months ago:
That’s just what they want you to think.
- Comment on America's Smartest Man Finds Something Interesting 3 months ago:
I’m still hoping this happens and leads to a WWE style outcome.
Elon has a heart attack on the ring and falls on top of Zuck pinning and smothering him. Zuck is forced to tap and Elon is stretchered out the ring
- Comment on Sorry 3 months ago:
Except that in industry no one gets tenure.
- Comment on Ambulances called to Amazon’s UK warehouses 1,400 times in five years 4 months ago:
Amazon are also dicks about sick leave. I’m sure forcing people to work, and work hard, when they’re ill leads to complications.
- Comment on Why are people downvoting the MediaBiasFactChecker not? 4 months ago:
Yeah anything responsible that reports and corrects mistakes made in past reporting get a low score for factual accuracy, when these should be getting the highest scores.
It’s completely backwards.
- Comment on Ya girl going in a Q1 4 months ago:
No it doesn’t work.
But it’s better than not doing it.
People suspect who the author is but maybe you cited those papers because you’re afraid of getting the author to review them, or you’re a fan-boying grad student.
- Comment on Celery 5 months ago:
No, it’s just that I actually work in the field.
I’m not describing binary classification, I’m describing multiclass. “Group classification” isn’t really a thing. Yes, your ml system probably guesses what kind of plant it is and then looks up the ediblity of components.
The problem with this is how they will handle rare plants that aren’t in the dataset, or that are in the dataset but with insufficient data to be recognised.
Because multiclass assumes that it’s seen representative data on all possible outputs (e.g. plant types) it will tend to be dangerously confident on plant types it hasn’t seen before.
This is because it can rule out other classes. E.g. if you’re trying to classify as rose, tulip, or daisy and you get a bramble, your classifier is likely to be very certain it’s a rose because tulips and daisies don’t have thorns. So your softmax score is likely to show heavy confidence in rose even though it’s actually none of them.
This is exactly what can go wrong when you try to use the softmax approach and come across an interesting rare mushroom or wild carrot. You don’t want it to guess which type of plant in the database it’s most like, you want it to say that it genuinely doesn’t know and you shouldn’t eat it.
- Comment on Celery 5 months ago:
The key issue here is that ‘level of certainty’ doesn’t really mean what you would like it to.
You get back a number yes, but it can change according to what’s visible in the background, the angle that the plants at, how close is it to the camera, and how nice the camera is you’re using (professional photographers use expensive cameras and take shots of different things to everyone else).
Interpreting this score as “how safe is it to eat the plant” is a really bad idea. You will still eat the wrong plant. These scores can lead to very confident random guessing when you show it a plant it’s never seen before.
And no, softmax is a trick for making the scores all sum to one, so you get back a confidence for every possible thing the image could be of.
- Comment on Do 9-5 jobs still exist in the U.S.? 6 months ago:
Teams defaults are pure scummery.
No, don’t alert me on a Sunday night with notifications that I might have missed over the last two days.
- Comment on Finish him. 🪓 6 months ago:
It’s worth saying that ml is in a very different position to most of academic publishing.
All of the serious journals are free to punish and fully open access and a significant amount of publication includes enough code that things are mostly replicable. GitHub has done wonders for our field. Also many tech companies use publications as an indication of prestige and go out of their way to publish stuff.
We’re still drowning in too many papers and 95% of everything is shit, but that’s every field really. Talking to musk on twitter is the not right place for a nuanced discussion about publication.
- Comment on Anon figures out how dieting works 6 months ago:
Doesn’t matter if you forgot to work out for a bit. The trick is to just start again when you realise you’ve stopped.
- Comment on Old comic, more relevant than ever 7 months ago:
The term machine learning was coined in 1959 by Arthur Samuel, an IBM employee and pioneer in the field of computer gaming and artificial intelligence.[9][10] The synonym self-teaching computers was also used in this time period.[11][12]
en.m.wikipedia.org/wiki/Machine_learning
It wasn’t so much stolen as taken back.
- Comment on Chatbot letdown: Hype hits rocky reality 8 months ago:
It’s good for anything that has thousands of examples on stack overflow.
For example, every time I end up trying to work with pandas, I always forget the syntax and it’s generally good here.
Anything unusual, or that is sufficiently complicated that I wouldn’t be able to Google for, and just forget it.
- Comment on Anon gets a job 11 months ago:
Having freedom and drive to do whatever you want is great, even if that means you spend a bunch of time just fucking about and getting high and watching TV.
On the other hand, a lot of the people that are dropping out like this are actually just depressed. They look like they’re doing the same thing, but they’re actually just self medicating and it sucks for them.
For some of these people, getting up and out the house, being forced to do a bit of exercise, and talk to people can help with minor depression.