hooande

21 ms·

181.
by hooande 5y ago
who would choose to see unengaging content? "My algorithm is set to show bland agreement with last week's news"
182.
by hooande 5y ago
This is a terrible comparison. With OSS people are looking for security flaws . With a twitter/google algorithm people are trying to exploit the system when it works as intended. If there is a secret formula for a successful tweet, pe
183.
by hooande 5y ago
the OP's point was that public reaction will depend on the specific joke you make and the context in which you make it. a skilled comedian can make a nonbinary joke that is perfectly acceptable. a person at random is more likely to hav
184.
by hooande 5y ago
He could just give them money for it. I'm sure they'd love having a multibillion dollar customer. It's very possible that his AI investments have yielded developments that we can't imagine. But it's hard to see the
185.
by hooande 5y ago
This is 4chan. Your recurring fantasy is 4chan
186.
by hooande 5y ago
As is detailed by many smart people below, fiduciary duty lawsuits are difficult to win. They'd pretty much have to have board members emailing each other saying "This would be in the best interest of shareholders, but let's
187.
by hooande 5y ago
people will immediately reverse engineer it to find what criteria give tweets the most engagement. if it's something like "joke tweet at 7 am" then millions of people will do whatever that is. Not to mention it will open it u
188.
by hooande 5y ago
every suggestion he's floated so far is bad. edit button, open source algorithm, "time outs" instead of bans, "free speech". all of these ideas have obvious flaws that he does not address and none of them would be s
189.
by hooande 5y ago
It's a poison pill. My short version understanding is that if anyone tries to control 51%, they can issue more shares to make it too expensive. There may be some kind of WallStreetBets technical maneuver around it. But it doesn't
190.
by hooande 5y ago
My general thought is that pump and dump schemes are small change to the richest person in the world. Making an extra few hundred million is appealing to anyone, but he'll make 100x that from TLSA by doing nothing at all. I do see this
191.
by hooande 5y ago
People seem to believe that it's impossible to create a competing platform. That's built on the assumption that most people are generally happy with current offerings. If Musk does anything that a critical mass of users finds obje
192.
by hooande 5y ago
if the algorithm is open source, bad actors can see it as well. it'll become remarkably easy to spam and game the system. this could drive users away, decrease revenue and leave Mr. Musk holding a $40B hot potato
193.
by hooande 5y ago
Say the board has a reason to believe that Twitter is going to double in value in the next two years. ie, they have a secret "Twitter Gold" product that they are certain will triple profits. It would be a disservice to me, as an i
194.
by hooande 5y ago
yeah, The Boring Company is absolutely killing it. I use that thing every day
195.
by hooande 5y ago
I get the reasoning of "general intelligence exists in humans, therefore it must be possible to create it". But this logic really applies to human artifacts and creations. ie, if a human builds a device that can do X then another
196.
by hooande 5y ago
> * Calculus for solving all sorts of problems. I'm not sure if I've ever used calculus thinking for anything as a full stack developer. Never done much with calculating the rate of change of a rate of change or determining the
197.
by hooande 5y ago
The problem with testing isn't that it's inaccurate, but that it's a poor tool for the problem that it's being used to solve. You can't express "likelihood of success in college" in one number. Looking bac
198.
by hooande 5y ago
Do you honestly believe that if you were from a very poor town, poor by the standards of african american neighborhoods in Detroit, that giving your child a $35 practice book would have a similar outcome?
199.
by hooande 5y ago
If there is a large shift toward gig economy work, then the healthcare industry will shift as well. There's nothing preventing them from offering affordable healthcare to gig employees other than the status quo and their current profit
200.
by hooande 5y ago
This doesn't make sense. There are countless opportunities for social interaction. Join a church, a club, a sports league, a gaming discord. I get that there's a lot of friction involved in joining anything, and many 'lonely
201.
by hooande 5y ago
1,000 words is much more challenging than, say 4,000. It's also limiting in terms of several of your stated criteria. You get a certain kind of story at short length. Generally you can get a higher level of literature around 3,000-4,00
202.
by hooande 5y ago
My understanding of flat earth was that it's about coming up with convincing arguments for something that you know (or at least assume) to be false. There are a large number of people who participate and they've been doing it for
203.
by hooande 5y ago
I like Gary Marcus as a personality and I look out for his work. Recently he's been doing this thing where he lists examples of deep learning failures (which are trivially easy to find) and then proposing symbolic / causal learnin
204.
by hooande 5y ago
This is like saying "I cured my alcoholism by throwing all the liquor out of my house". The problem isn't the availability of the stimulus, it's the lack of control over your desire for it. You can't cure a need for
205.
by hooande 5y ago
Disney has been doing something similar for years: https://superparent.com/article/276/osmo-and-disney-make-kid... I remember an article that they had something more general (not just Frozen) available for childre
206.
by hooande 5y ago
Not sure why you're making a distinction between ML and deep learning here. both of them can be black boxes. Calculating the area of a square can be a black box if all you know is how to plug numbers into the formula. A big part of mac
207.
by hooande 5y ago
This isn't surprising. Most 'AI' comes down to correlating a small number of variables (one or two) with a prediction target. The real benefit of any form of machine learning is detecting functional relationships between vari
208.
by hooande 5y ago
great app, thanks! glad that a skilled hacker could solve this problem
209.
by hooande 5y ago
that money was never real, they never had a chance of spending even 0.1% of it. lol does anyone really think they could have accessed and spent that much money, in any way?
210.
by hooande 5y ago
The takeaway here seems to be that black inventors in the northern United States, where there was ostensibly less racism, earned patents at nearly the same rate as every other race. Black inventors in the south, where the racism was extreme

More ›