I agree with you. Replication is key for validating the claims made by one person/team. However, too much of it draws resources that could be used to fund new research. How much is too much? I don't think there is a precise answer to that, but I guess that when the research is very expensive and/or depends on some rather unique circumstances, I would be inclined to avoid extensive replication in favor of more scrutiny over the data collection and analysis.
As in almost everything, the dose makes the poison.
This guy is absurdly shortsighted. How can IBM's board allow someone with so little vision run a technology company?
Maybe the board is pushing in the same direction?
A very close friend of mine, founder and CEO of a successful startup, tried to make his company almost 100 percent remote after the pandemic. This move made complete sense as they are in a niche online retail sector that is, in a nutshell, a website coupled with a well-oiled decentrailzed logistic operation.
The board refused, forcing his hand, which is stupid. They pay extra for prime real estate that also brings in more risk, as the internet connection is concentrated in one location. No matter how much redundancy you have, you can't beat having people distributed in many locations (as it would be if they were allowed to work from home), given that they can do their jobs independently. Also, the company is losing valuable employees that are no longer willing to spend two hours per day in traffic jams.
Yes, Mr Krishna is pushing a dumb move, but that doesn't mean he is the only stupid guy up there.
First, "culture eats strategy for breakfast". This is a famous quote by Peter Drucker that implies that no matter how elegant a strategy is (and how sound the predictions on which it is based are), if it goes agaisnt the cultural dynamics, it will fails catastrrophically. Does anyone really believe that most parents or schools will decide to outsource reading lessons to a corporate owned machine/software, in a year and a half from now? Unless, of course, the prediction is about a couple of isolated cases that, althoug they will barely count as evidence, will be touted as the next coming of Jesus by our corporate overlords.
Also, asuming that the objective is to get kids to read well (and read more), then the Ai should encourage parents to read and buy books. The single most significatn factor that predicts success foir kids in many countries is the amount of books they have at home. Not childrens books, just any book will do. That's because the reason they learn to read well and they do it often is because their parents do!
This is what a high schoo chem teacher actually said, implying that chemical reactions run the world. Of course, in a reductionist view this makes sense, like it will also for physics, math, language, etc, depending on your point of view (and mostly on your background).
Of course, "disciplinary reductionism" is tempting and a main source for most claims of "X is essential, we should teach it to everyone". But there is an additional underlying assumption that gives rise to these claims: individualism. In a way, these proposals run on the desire to have everyone to be able to function with as little reliance on others as possible. They are a product of mistrust.
I don't need to know about cybersecurity, like i also don't need to know how to physically secure my home. There are people that know more than me about it and I can ask them and follow their recommendations. It is certainly relevant to know that there are such things as cyber threats, but there is a difference between knowing something exists (and identify it) and knowing what to do when you encounter such a thing.
Going back to the High School curriculum, we should encourage kids to learn the existence and relevance of as many things as possible. Doing that requires a curatory effort to balance the need for giving each topic a minimum of class time with the limited amount of total hours and effort that hte students can devote to learning. But this balance usually means not to prioritize any specific competence. After elementary and middle school, when children already have the basic competences for interacting with others (language, math, group work, play, sports, etc.), the goal is to open the space of possibilities for each student, not to direct them towards whatever fits the current fad being touted in the media.
"Choosing scientists to be administrators based on sex seems like it might have some pitfalls.
Well, the fact that 100% of her predecessors were male could also be seen as a result of a process that was incredibly gender biased.
I do wish they would erase the hobbit trilogy and remake it, though.
100 times this. I am a Tolkien fan and loved the LOTR movies. But the Hobbit trilogy is so bad that I coudn't watch the 2nd and 3rd movies even when they were released on streaming/cable. It is a shame to bring down such a wonderful story to a cliché-infested, blockbuster kind of movie.
Too bad the USA is not part of the American Convention on Human Rights, therefore it is not under the jurisdiction of the Inter-American Court of Human Rights. Citizens of member states have the right to present their cases when there is no recourse in their respective countries.
This is the same thing the IMF has been doing for decades and it has been VERY wrong except a couple of times (you know that even a broken clock tells the correct time twice a day).
This Financial Times article is literally titled "IMF shows poor track record at forecasting recessions". Between 1991 and 2017, their average prediction was 5, meaning that five countires will be in recession the following year. The actual average was 26... If this is not a total failure, I don't know what would qualify as such
I see that their current prediction differs from their past trend (it is much higher), but unless they show some real improvement, backed with accurate predictions, then you should not take them seriously.
"A car is just a big purse on wheels." -- Johanna Reynolds
