I am not sure that 400k subscribers translates to 400k paying users. As a student I may use it for free for example.
40 karma · joined June 28, 2022
Given the scientific evidence from fMRI, and post-partum biopsies of their brain, those things are very much real. They do not misinterpret reality.
Schizophrenic people do misinterpret reality, whereas transsexuals do not. Had transsexuals misinterpreted reality, they would not experience GID or Gender Dysphoria, nor would reality ie their sex, be a stressor.
> We should be very clear about the results of transitioning.
People are very clear about the results and consequences of transitioning, people don't do it on a whim, and fyi, transitioning is a very long and arduous process. It does not happen over a day; over a week; a month; or even a single year. The person chooses to walk that path every single day, and at least in Europe, they are well aware of both the consequences and the results.
> It does not result in female anatomy by an regular definition. It is a physical stake that imitates the <female/male> anatomy, very superficially, with long term health consequences and perpetual maintenance
Nobody is claiming it provides the person with a uterus, ovaries, or - if the person has transitioned into adulthood - skeletal structure.
On the other hand, it has very real changes; from brain region activations, taste, muscle tissue change, skin change, and a lot more. To claim that it does nothing would imply that it is needless, had it been needless, people wouldn't be fighting for better coverage, wouldn't ask for treatment, and on and on.
> I suspect that this terminal state does not achieve the goals of trans folks,
Sorry but at this point you are imposing your beliefs to argue for the experiences of other people, millions of them who contradict yours. At this point, it is good to listen carefully with the intent to learn over claiming what you believe is the truth.
> is likely a major source of the ongoing emotional disturbance so commonly observed.
Transition is a source of emotional disturbance when the person is not accepted, in fact, discrimination in workplace, housing, along with overall lack of acceptance from the broader society is the main source of stress and the main cause of detransition.
In-fact, people who transition early pass significantly easier, have the same outcomes as cis people, and get to live a happy life, in fact over 98% of people who transition early do not detransition. That's a big super-majority.
I recommend spending more time with actual trans people and being more empathetic towards people whose shoes you can not walk in.
For them to pull out it seems that associating with twitter harms their image, and that’s - according to them - related to moderation.
As for “proof” that twitter is more toxic;
> In the week leading up to Musk's acquisition, researchers said there were no more than 84 hostile tweets an hour on the platform. But from midnight on October 28 – the day Musk took ownership – to noon the next day, there were 4,778 hate-filled tweets. That accounts for more than 398 tweets an hour, about 4.7 times higher than the seven-day average leading up to that day. The potential reach for those tweets was more than 3 million, researchers found.
> Researchers also said there was an increase in negative sentiment, with more than 67% of the tweets sent after Musk's takeover having a negative tone.
https://www.cbsnews.com/news/elon-musk-twitter-caused-measur...
However, "reverse-mode" AD is not as simple as just caching results.
> BP’s continuous form was derived in the early 1960s (Kelley, 1960; Bryson, 1961; Bryson and Ho, 1969). Dreyfus (1962) published the elegant derivation of BP based on the chain rule only.
> explicit, efficient error backpropagation (BP) in arbitrary, discrete, possibly sparsely connected, NN-like networks
But that wasn't close enough to current BP
> BP’s modern efficient version for discrete sparse networks (including FORTRAN code) was published by Linnainmaa (1970). Here the complexity of computing the derivatives of the output error with respect to each weight is proportional to the number of weights. That’s the method still used today.
See [1] for more details.
I do not know the exact implementation details of Linnainmaa's BP, but there are all sorts of improvements to plain reverse mode auto-grad through stuff like Jacobian-Vector products that save an enormous amount of memory.
Nocedal's "Numerical Optimization" covers a great deal on how to implement autograd, and I can easily recommend it to anyone interested.
[1] https://www.reddit.com/r/MachineLearning/comments/e5vzun/d_j...
Time of publication. I merely pointed that Schmidhuber et al's results were published long before Hinton et al's work (I included year of publication for a reason), ergo the comment that was in support of Hinton was refuted and my comment gives further validity to Schmidhuber's claims of misattribution and plagiarism, while pointing at yet another paper tiger in the claims that the three pioneered the field.
If anything, Schmidhuber may have deserved a Turing award on his own, and what we see is the result of politics.
This is pure speculation, but to me the whole ordeal hints at manipulation from the companies where 2 of the 3 worked at, which would essentially increase the pool of candidates the respective companies have only by association with their employee.
According to [1], the papers [42,43] showed that plain backprop was capable of solving complex tasks without pretraining on a GPU.
From [1]:
> Our team then showed that good old backpropagation [A1] on GPUs (with training pattern distortions [42,43] but without any unsupervised pre-training)
I can't verify since I have a barely stable connection and little battery, but somebody else should be able to.
[1] https://people.idsia.ch/~juergen/firstdeeplearner.html
[42] H. Baird. Document image defect models. IAPR Workshop, Syntactic & Structural Pattern Recognition, p 38-46, 1990
[43] P. Y. Simard, D. Steinkraus, J.C. Platt. Best Practices for Convolutional Neural Networks Applied to Visual Document Analysis. ICDAR 2003, p 958-962, 2003.
- Spotlight, i.e. the search bar, and
- Preview, i.e. the image/pdf/document viewer and editor. It's significantly more feature rich than most pdf readers, allows editing pictures with significantly more tools than paint, allows you to add a signature to sign documents. Absolutely great piece of software.
As others have mentioned, similar stuff exist in other countries as well.
So to me, I'd rather have a single common, shared and opensource system for both Denmark, and all of EU.
One of the interviews today was focused on the task that I will be working with, we started from a simple instance of the problem, and saw how we could improve upon it. The task revolved around querying and identifying entities. We worked through multiple instances to see how we can improve on the initial approach, or how to handle odd cases.
After that some leetcode was included.
The other two interviews were more on general technical and ml stuff.
I found this approach to better than other kinds of interviews I have gone through.