geraneum

16 ms·

331.
by geraneum 3y ago
It’s not that straightforward. Quoting Yann LeCun from a LinkedIn post [1]: > I have claimed that Auto-Regressive LLMs are exponentially diverging diffusion processes. Here is the argument: Let e be the probability that any generated tok
332.
by geraneum 3y ago
We still call bitter oranges “Narenj” in Persian. That’s the root of the word Orange as the author puts it. That one hasn’t changed much.
333.
by geraneum 3y ago
Go was first and foremost created to serve the needs of Google. Kudos to them for releasing, open sourcing and backing it but it was designed to serve their needs. It’s not a criticism but it explains why it’s quirky in many areas. The auth
334.
by geraneum 3y ago
“autoclaves, antibiotics, and anesthesia during surgery are nice but the super intelligent AI that found the cure for all illnesses and the ability to go back in time and etc. are more important.” Someone in x thousand years from now.
335.
by geraneum 3y ago
Obviously, compared to other means of transporting water at that time! For example an “open” river could have been more vulnerable in certain circumstances. Strong earthquakes may affect the path of the rivers as well. But I see your point
336.
by geraneum 3y ago
A little side note, the original name of Qanat in Persian is “Kariz”.
337.
by geraneum 3y ago
“You will drink nothing behind the wheel until we approve it.” “You will wear this belt because we say so. You have no freedom.” If _they_ don’t want you to go somewhere, your car surely won’t take you there.
338.
by geraneum 3y ago
> to ensure it never does something But there are many systems for which you cannot predict/control the behavior with just a few experiments because they are simply, probabilistic. Isn’t it also the case with LLMs? If not, why?
339.
by geraneum 3y ago
> You can make it more likely to refuse a request but the idea that you can guarantee it won't is completely wrong. Why is that the case, technically?
340.
by geraneum 3y ago
Sometimes (more often than people realize) computers roll a die and proceed accordingly.
341.
by geraneum 3y ago
You are actually right. You didn’t claim that the Greek did it first and that is what I doubted in the first place. I misread your comment. That fact that the Greek could prove the theorem at some point is not unlikely even without evidence
342.
by geraneum 3y ago
From the article: > Pythagoras is immortally linked to the discovery and proof of a theorem that bears his name – even though there is no evidence of his discovering and/or proving the theorem. Simply quoting. It seems like > Gre
343.
by geraneum 3y ago
There was a time when people from Finnland were discriminated against in US as they were not considered “white” because of their language being of Uralic origins and not Indo-European. That’s how ridiculous it was/is. https:/&#x2
344.
by geraneum 3y ago
In Persian it’s “na”. Also the words for Mother, Father, Daughter, Brother and some other primitive concepts are very similar. But it’s not really surprising.
345.
by geraneum 3y ago
IP addresses are considered personal data according to [1] and there’s no mention of it being counted as personal data conditionally. [1] https://commission.europa.eu/law/law-topic/data-protection/r...
346.
by geraneum 3y ago
It was said that after the 9€ ticket experiment, they wanted to analyzed the performance and outcome of such a scheme and I _assume_ that this has been somehow calculated to not over burden the system. I have a 49€ ticket (company provided)
347.
by geraneum 3y ago
What is a free passenger?
348.
by geraneum 3y ago
This is our natural tendency, to seek the quick rewards but we are intelligent beings, capable of planning. The same arguments can be made for future of your own children, retirement plans, planting trees, long term investments, etc. Yet th
349.
by geraneum 3y ago
> Yes ahead of Uighurs in China, Women in Iran and Afghanistan. When they said these are important matters in TV, did you really believe them? These, among many other issues, are not even in the queue my friend.
350.
by geraneum 3y ago
I doubt the same people who won’t accept a decline in quality of life by reduced oil consumption will be thrilled with the consequences of the alternative.
351.
by geraneum 3y ago
From the article: > Corrado said that while he wouldn’t want it to be a part of his own family’s “healthcare journey,” he believes Med-PaLM 2 “takes the places in healthcare where AI can be beneficial and expands them by 10-fold.” Hah?
352.
by geraneum 3y ago
Does it mean that if my thermometer shows 10C it might be lying because of its political bias? Or did you miss an “/s” and I the sarcasm?
353.
by geraneum 3y ago
Yes, I agree. It’s very unlikely that humans disrupt the natural habitat of other species.
354.
by geraneum 3y ago
> A bit like watching a (good) silent movie. A bit off topic but in this context I found “Dogville” to be interesting.
355.
by geraneum 3y ago
While what you mention (capacity) is an absolutely important issue, I am skeptical of the effect that those 1-2 kids could have had on the students' performance on a national level so suddenly. Immigration is not a new phenomenon here.
356.
by geraneum 3y ago
The problem is having an immigration background in Germany doesn’t not mean the students don’t speak German [1]. Even if their parents came from Turkey or Russia 50 years ago, they are classified as such. It also doesn’t explain the deficie
357.
by geraneum 3y ago
It say in schools “where” 70 to 90 percent have immigrant background. I don’t think it means that portion of all students everywhere on average have immigration backgrounds.
358.
by geraneum 3y ago
I am curious what makes it absolutely insane in your view?
359.
by geraneum 3y ago
The compilers _usually_ generate correct machine instructions (code) better than us. This, alone, is not really a meaningful measure and doesn’t refute the claim.
360.
by geraneum 3y ago
Ah this came to my mind as well. Although I didn’t consider it a serious thought. But then if one day we find out that this was the purpose (for a similar service), I won’t be surprised.

More ›