This inflicted pain and damage appears to have no significant impact on adult quality of life though. Circumcision may also even reduce the bacterial load and hence reduce chances of infections and cancer: https://slatestarcodex.com/2019/12/10/acc-is-infant-circumci...
It's a form of genital mutilation. It would be wrong to call it a non-issue. It is moral one at the least.
I have no issue with consenting adults getting circumcised for what ever reason they want too. A baby, however, cannot possibly consent to such a thing.
> A baby, however, cannot possibly consent to such a thing.
Well, a baby cannot consent to anything. The parent's consent is the baby's consent.
Now I'm genuinely curious - I'm not affected, so I can't relate: Circumcision is genital mutilation, sure, but is it really "that" bad? Does it affect quality of life, from child to adult (excluding botched circumcisions)? Based on my understanding, I would rather focus on female genital mutilation in undeveloped parts of the world. In comparison, male circumcision seems to be ... ok. It's fine. Or is it? Please let me know.
(I'm not talking about the ignorance of infant's sense of pain here, that's clearly immoral).
What matters the most is the cohesion, functioning and productivity of society, which is unaffected by circumcision.
The notion of consent is arbitrary. You could also regard religious cults as primary, then you would need their consent to imposing your personal aesthetics.
Smashing one in a hundred babies to death with a frying pan wouldn't impair the functioning or productivity of society either. Seems like a pretty poor way to measure barbarism to me.
Cutting the outer ears off an infant would also have no significant impact on adult quality of life, yet if a parent tried to do that they'd be arrested.
As someone who didn't face circumcision complications until puberty, I'd like to track down the doctor who did it to me and see how he likes having pieces carved off of him.
So the idea is that not consented mutilation is defensible as long as the consequences after a few days or weeks are extremely minor? Just repeating this back to you so we're on the same page.
I'd say that attempting to prohibit this tradition has a net negative impact on the integrity and functioning of society because the harm in question is very minor but the reaction to such an attempt will be major. I'm sure you are trying to twist this into "but then I could anesthetize and then circumcise you", however, in that case I'd retaliate so drastically that it would negatively impact the functioning of society much more than say 100 circumcisions, hypothetically speaking.
I’m not trying to twist this into anything, I literally only repeated your own words back to you. So now this isn’t about whether the consequences are minor, but whether the person being harmed has the power to retaliate and how such retaliation impacts society? I’m having a really hard time following your rationale, and if I can be perfectly honest the impression I get is that you first decided that genital mutilations in infants is fine and are now retroactively looking for justifications.
Power is one factor informing policy, empathy is another. Neither is sufficient for a policy that sustains the system for long on its own. Only focusing on power would result in e.g. exploitation of Africa. But most don't like this due to empathy.
Focusing only on empathy often misses the point too as it tends to get politicized, overinflating minor issues as in this case, which will ultimately result in power conflicts, degrading long-term sustenance.
Lilium's plane has 30 engines, multiple independent battery packs, and electric motors are way more reliable than jet engines (due to being so much simpler). A Lilium plane may have a 1000x higher chance of a crash in case of full engine failure, but perhaps that can be compensated for by 1000x times simpler/redundant propulsion technology. This seems to be very hard to assess in theory, so we probably just have to wait and see how it performs in practice.
That's all very well, as long as there's no common modes of failure. Redundancy can be a means to achieve reliability, but not always. Imagine the cause of failure is a bug which kicks in at a specific time of day, or an integer overflow which happens after a certain amount of uptime, or all the motor drivers are susceptible to a specific RF frequency. It doesn't matter how many motors you have if they're all susceptible to the same failure mode.
> integer overflow which happens after a certain amount of uptime
It’s not just electric powered planes that have to worry about such things… this was an issue for the Boeing 777 Dreamliner too. If it was powered on for longer than 248 days, it could lose all electrical power due to an overflow in the generator.
I’m not saying it isn’t a concern, but rather it is a concern for all planes (and vehicles for that matter). Many commercial passenger planes are now fly by wire. If you lose electrical power, you’ll also lose control. So, while we’re talking about purely electric planes, the problems are universal.
> 1000x times simpler/redundant propulsion technology
This is a contradiction.
A system with that level of redundancy will have a corresponding increase in complexity of management systems. That means software (multiple copies of software on multiple independently powered computers all somehow coordinating). This makes the management software the single point of failure, and frankly I'd sooner trust a 50 year old pair of mechanical engines to a 5 year million line of code program.
You can always build things more safely with more resources (more maintenance, more checks, more testing etc.), the question at the end of the day is whether it is economical. I don't think this question can be answered upfront in this case.
The model only has about 1B parameters which is relatively small.
The language models that produced very impressive results have >>50B parameters, e.g. GPT-3 with 175B, Aleph Alpha Luminous (200B), Google PaLM (540B). GPT-3 can understand and answer basic trivia questions, and impressively mimic various writing styles, but it fails at basic arithmetic. PaLM can do basic arithmetic much better and explain Jokes. Dall-E 2 (specialized on image generation) has 3.5B parameters for the image generation alone and it uses a 15B language model to read in text (a version of GPT-3).
This proves that none of these institutions and politicians pushing for more surveillance actually care about the problem (while the urgency of the problem is dubious to begin with).
It's all about power and funding. If the problem was eradicated, then they would lose their purpose and funding.
Do you really pay for less? Digital signage displays often run 24/7 in temper proof cases with relatively poor ventilation and heat dissipation. Thus you might actually get a display that lasts longer (which, AFAIK, mostly comes down to higher quality capacitors and capacitors placed where they do not overheat, e.g. not directly next to high current transistors and resistors, which is common planned obsolesce practice in consumer TVs).
That's not entirely true. Neural networks are fairly robust to noisy training data (a.k.a. garbage).[0] Well, stochastic gradient descent has the noise in its name. More training data can compensate for noisy data to some extent.[1] I'm not sure know if model size can also compensate for noisy data though, but would not be surprised if it did.
There are very specific conditions for this to hold, mostly that the incorrect sample is surrounded by correct ones, and that the model is small enough or the error vanishingly rare. Notably the reference you gave also shows horrendous generalization performance, so its really just showing how easy it is to overparametrize. Input errors can be accounted for to some extent, but also under specific circumstances, eg.
I mean the argument there is basically if there is enough good quality data, the bad data is somewhat (or mostly) compensated for. To which I would argue that is no longer a “garbage in garbage out” situation as most people use it.
It's a non-issue.