AI Can Now Help Write Wikipedia Pages For Overlooked Scientists (popsci.com) 97
An anonymous reader quotes a report from Popular Science: Plenty of prominent scientists have Wikipedia pages. But while checking to see if someone specific has a Wikipedia page is a quick Google search away, figuring out who should be on Wikipedia but isn't -- and then writing an entry for him or her -- is much trickier. For example, you may or may not have heard of Christina Economos. She doesn't have a Wikipedia page about her, although she's a professor at Tufts University and the New Balance Chair in childhood nutrition. But while she lacks a Wikipedia page, she does have a very short stub describing who she is professionally on a website made by a company called Primer. That little blurb, which could one day grow into a full-blown Wiki entry, was created by an AI system dubbed Quicksilver. The idea behind the project is to use AI as a jumping off point. Humans can use it to help them write Wikipedia pages for scientists who don't have them, but deserve to. For example, on Economos' Primer page, there's a link to an article from CBS Boston that mentions her -- a good potential source for a human Wikipedia editor who may want to write an entry for her.
Primer launched officially last year and uses AI to read information and generate reports; part of its focus is doing the kind of work an intelligence analyst might do. Artificial intelligence generally needs data to learn from, and so for this project, Primer used around 30,000 existing scientist Wikipedia pages to train their machine learning systems. Then they fed 200,000 names and related employment information into their AI system. Those names came from the listed authors of scientific papers focused on computer science and biomedical research provided to Primer from the Allen Institute for Artificial Intelligence. If you're curious to see a sample, you can head on over to this page, which has 100 examples of AI-generated Wikipedia blurbs.
Primer launched officially last year and uses AI to read information and generate reports; part of its focus is doing the kind of work an intelligence analyst might do. Artificial intelligence generally needs data to learn from, and so for this project, Primer used around 30,000 existing scientist Wikipedia pages to train their machine learning systems. Then they fed 200,000 names and related employment information into their AI system. Those names came from the listed authors of scientific papers focused on computer science and biomedical research provided to Primer from the Allen Institute for Artificial Intelligence. If you're curious to see a sample, you can head on over to this page, which has 100 examples of AI-generated Wikipedia blurbs.
Re: Welcome to the circlejerk (Score:1)
But the AI have passion for their work...thAI have a purpose!!!! Dammit... they deal with people, who in turn treat them like some dusty computer cloud. Colosus forever!!! Wait...Trump won so Guardian forever!!!
How sad (Score:5, Insightful)
Little known scientists need an AI to write a Wikipedia entry about them. Yet there are plenty of humans interested in creating Wikipedia pages about any minor sports personality in all languages. Here for instance, I searched an obcure Belgian soccer player in the Finnish version of Wikipedia and found it [wikipedia.org].
Sport is clearly more important than science it would seem...
Re: (Score:1)
Re: (Score:2)
Sport is clearly more important than science it would seem...
No, not more important, just more popular. But the solution is more articles about scientists, not fewer about sports. It is not a paper encyclopedia, and there is no inherent limit to the number of articles.
Re: (Score:2)
Sport is clearly more important than science it would seem...
No, not more important, just more popular. But the solution is more articles about scientists, not fewer about sports. It is not a paper encyclopedia, and there is no inherent limit to the number of articles.
Going to the PopSci article, this is apparently a gender issue - which brings up some other issues.
I haven't yet seen mention any mention whether or not the woman has given consent to someone or something writing a Wikipedia article about her. I can see some real trouble brewing there.
What if she doesn't want a Wikipedia page? What if she disagrees with the Wikipedia article?
Aside from the article's intent to focus on women, its obvious the same issues can apply to men.
I dunno, a sexist Ai bot writi
Re: (Score:2)
Did Marie Curie give consent for her page?
She dead.
Anyhow - here is some sit down truth for you.
Marie Curie is famous not because the is a female. She is famous for what she accomplished. Until the present day, that is
Do you understand the difference? This AI bot that writes Wiki articles is stated to give women more exposure. It isn't necessarily because they are doing world class work, it is because of their sex.
Marie Curie said "I'm gonna find out what the ray emitting substance is in pitchblende - hold my Beer!"
And she collected 2 well
Re: (Score:2)
Poor example, as in 1903 Mrs. Curie was originally NOT going to be awarded the prize as she was a woman.
And what about Rosalind Franklin?
Rosalind Franklin - she did not receive a Nobel Prize for a very good reason. She was dead.
In 1958, when she died, DNA structure was not yet proven. After a few more years, in 1962, the Nobel was awarded to Crick, Watson and Wilkins for their body of work on nucleic acid.
Crick had even suggested that Franklin be awarded a Nobel in Chemistry. However, Nobel rules prohibit posthumous awards.
So the good doctor Franklin has been awarded many other honors for her work, including a University bearing her
Re: (Score:2)
The rule preventing posthumous awards was introduced in 1974. Crick et al received the prize in 1962. What was your point again?
Doen't matter. Of the 923 Awardees, there are only 2 Awards given out after they were dead. They hate giving awards to dead people. They just made it official.
And apparently you missed everything else > Why don't you read and refute that? I'll wait.
Re: (Score:2)
. Feminists have reasonable cause to feel aggrieved. But then anyone interested in science or who believes in a meritocracy should also feel aggrieved. Points should make prizes, not penises, unless you can show me that you have to use your penis to hold a test tube over a bunsen burner.
Well now - you are a pretty disgusting sexist there, Well considered that you post as an AC.
I am pretty certain that the only cure you would accept is for a special division of Nobel prizes exclusively for women.
At that point, your utter fixation on the specific genitalia a person is wielding might be assuaged.
After that we can consider Nobel prizes only for gay people, trans people (would male to female and female to male be the rule, or would one encompassing group be okay?)
A meritocracy based on wha
Re: (Score:3)
Subjects generally don't get a say in if there is a Wikipedia article about them. There are stricter rules for biographies of living persons, but the existence of an article is governed by the rules on notability rather than the desires of the subject.
Otherwise people whose articles say unflattering things about them, particularly politicians, would get them removed.
Also, if someone were to ask someone else to write an article about them, that would contravene the rules too.
Re: (Score:2)
"What if she doesn't want a Wikipedia page?"
If you are famous enough, you get a Wikipedia page like it or not. There certainly are people who wish they didn't have one.
If you are not very famous and complain, you can get your page deleted.
No argument there.
I have also written pages for 3 people (who did not ask me, but I asked them) and maintain one.
Which is why posting anon.
Oooh! could you write an Ol Olsoc page?
Re: (Score:1)
To be fair, Romelo Lukaku is not obscure at all and is one of the best players the Belgian team has ever had. He was often seen as one of Europe's ten best players.
In Europe, soccer is big.
Re: How sad (Score:2)
People just are not as interested in obscure researchers who may not have done any ground-breaking work or have an extensive career behind them.
This just seems pointless and will probably end up with outdated information (since no one cares to make the page to start with) and will li
Re: (Score:3)
This just seems pointless and will probably end up with outdated information (since no one cares to make the page to start with) and will likely get vandalized by mischievous students.
After reading the article,it is a political thing. The point is to "increase the representation of women scientists on Wikipedia".
But seriously, this is all being done without consent, so starts off on shaky ground.
Is the AI article writer sufficiently non-male biased?
Should the AI article writer be precluded from writing about males?
Transgender scientists?
What about names that are not easily assigned to a sex - we don't want to assume gender.
Silly? You betchya. But if you introduce gender ident
Re: (Score:2)
Not even necessary to read the whole article to come to that conclusion. After I read blah who should be (vulgo deserves here comes the white knighting) on Wikipedia it was a dead giveaway that this smells of politics of identity politics to be precise. Wah wah women deserve to be treated equally. Women are as good wah wah. Excuse me but who wants to research the ratio of male to female scientists that are active at the moment. Or overall. Ok is guess that gets hard to find out but I bet the ratio is similar to the ratio of male to female scientists listed on Wikipedia.
The female scientists and engineers I worked with hated the identity politics terribly. Getting trotted out for talks when visitors came through because it was suppposed to promote women. As one said to me. "I wanted to be an engineer, and I've become a nerd fashion model."
I think that so many identity politics people don't understand that women who become scientists can't stand being looked at a sex objects by the feminists. And yes, if a woman is looked at as a woman before being looked at as a scientis
Re: (Score:2)
Gender identity politics was very much present in science in the past, and led to a failure to recognise female scientists.
Of that, there is no doubt. But eventually, we have to move on. the best way to eliminate 'isms is to do good work.
The horse has bolted the stable, and has probably led a full life and gone to the knacker's yard by now. Thankfully there is now less discrimination.
The problem with activist groups is that they tend to want to remain activist. After achieving their goals, they tend to overreach, or move the goalposts. Mothers against Drunk Driving is an example. After quite a bit of success in getting legislation passed, they continued, to the point of a BAC that isn't drunk is considered DUI. Around here, after the new laws were passed, MADD started spor
Re: (Score:2)
Your argument practically writes itself:
Wikipedia lacks articles about important subjects because Wikipedia itself isn't important enough to have them (contributions accruing only from editors "without anything better to do").
This is the most juvenile behaviour known to the human species: pointing to something you find unimportant and screeching "why is th
sibling-envy boo birds (Score:2)
Belated subject line to my previous post: sibling-envy boo birds.
Re: How sad (Score:2)
Re: (Score:1)
When did wikipedia become a place to post resumes? (Score:5, Insightful)
Re: (Score:2)
This. I predict a week later, these will get auto-deleted by a bot for being insufficiently noteworthy. And thus will begin the first AI-powered edit war.
It Serves Feminism, not Science or Wikipedia (Score:5, Interesting)
Another aspect of the project is to make it easier for scientists who are women to get the representation they deserve on Wikipedia—to empower human editors “to close the gender gap in representation of women in science,” Bohannon says. One of the ways that can happen is if a group wants to create more Wikipedia pages with a focus on women scientists, they could use data from Quicksilver, which Bohannon points out is filternable by gender.
This is yet another sexist politically-motivated project, not one that genuinely cares about scientific merit or improving Wikipedia.
Re: (Score:3, Insightful)
Equality, like feminism, has two modern meanings. The polite one when anyone asks what the definition is, and what is practiced. Both now seek a future where women, regardless of merit, are treated better than men in every context.
None of these efforts end even when women do better. There's still thousands of women-only scholarships and tens of thousands of programs to get more women to go to college and give them support when they are there despite being the majority of students and graduates by a large a
Re: (Score:1)
Someone creates an AI system that writes Wikipedia articles about people, regardless of gender.
They note that one potential use would be for groups interested in getting more notable women from the sciences on to Wikipedia to use these articles are the basis for one. One of the issues with Wikipedia is the lack of contributors, especially to less well known people's articles.
Somehow in your mind this makes the whole thing sexist and politically motivated. Most people would see it as an attempt to help impro
Way to lower Wikipedia's quality (Score:4, Interesting)
Now AI is going to trash Wikipedia with useless stub articles based on information you can google within 10 seconds That's just what we need.
Re: (Score:1)
Can vs Should (Score:2)
Sure they can, but should we let them?
automating what shouldn't be (Score:3)
I'm not a huge fan of Jimmie Wales, but one thing he said made a lot of sense to me-- he commented that at wikipedia they're continually at war with programmers who want to automate things that are better done by a human being... e.g. it's easy enough to send a standard welcome message to every newbie, but because it's a standard message it doesn't mean very much, and it's better to have a culture where actual human beings decide to send out welcome messages...
Automatically generating pages for subjects that a human being couldn't be bothered with sounds like an idea that is perhaps not quite as dumb as letting people vote by cellphoe, but it's getting there.
Re:automating what shouldn't be (Score:5, Interesting)
For some things, automatic pages are appropriate.
There is a guy who has "written" 2.7 million Wikipedia pages [wsj.com]. For example, he created a page for every single bird species where the pages don't already exist. That's OK because the basic information for each species is pretty formulaic - English name, Latin name, classification, habitat perhaps. Once the page exists, humans can add more "interesting" info if they have any.
This method doesn't work well for other topics, like people.
Re: (Score:2)
This method doesn't work well for other topics, like people.
Well it works for some people, for example if you say that everyone who's won an Olympic medal is a notable athlete, everyone who's won an Oscar is a notable actor, everyone that's won a Nobel prize is a notable... something, everyone in Congress is a notable politician and so on. Of course you're just then moving the discussion from the individuals to the qualifying criteria, because there's a lot of crappy competitions and awards and prizes and local politics where you're at most notable to a very limited
If a Wikipedia page pops up in the forest... (Score:2)
..is it visible?
So this AI bot creates a Wikipedia page ( barely a snippet from what I saw ) from already accessible information. Is there going to be another bot trolling Wikipedia to find pages to link to this page? How will it fold the link in context? Because if no other Wikipedia page links to it people won't find it within Wikipedia without explicitly searching for it. And since no pages link to it, and it contains the same content as the easily discoverable original page it's going to be low rank
More importantly... (Score:2)
Can AI delete Wikipedia pages for non-notable scientists who write themselves a page in order to promote themselves?
Re: (Score:1)
A few years ago I used to know a guy in college who was part of the RC Patrol on Wikipedia. He often would show me hilarious and dumb vandalism/trolling on Wikipedia articles.
One thing I always remember about it though is that they frequently had articles for various Indian guys who were just... office workers.? Maybe some minor programmers, but none of them had any kind of notable achievements. These articles put up daily about these guys whose names showed up pretty much nowhere on the Internet.
Just for scientists? (Score:2)
Why would this method be limited to scientists? Couldn't it be asked to write up a bio of anyone?
Re: (Score:1)
It was specifically trained to pull bio information out of research papers and most likely uses citation count or the text of articles that cite the author's work in order to determine the most influential topics written by the author.
I met many prominent scientists in my life... (Score:2)
Privacy? (Score:3)
Then they fed 200,000 names and related employment information into their AI system.
Before doing this, I sincerely hope Primer got written permission from those "overlooked" scientists.
One reason for not having a Wiki page is because they don't actually want one. Not everybody is a self-promoting narcissist.
Re: (Score:2)
Ah (Score:2)
How do you define 'deserves'. (Score:2)
I kind of thought the idea of a crowd-sourced encyclopedia was that the community decided which topics were important ( aka deserved) to be covered and then wrote about them.
So ok the AI can write a bio, but does a person decide who 'deserves' one? If you don't know anything about them how would you decide that? If you know something about them , why not write that?
Needs work but very cool (Score:1)
"Auriel A Willette is affiliated with Iowa State University.[1]He specializes in food science and human nutrition.[2]He is a member of Department of Food Science and Human Nutrition.[1]
Willette’s work focused on an area of the brain — the medial temporal lobe and specifically the hippocampus — that is critical for learning new things and sending information to long-term memory.[3]She and Webb analyzed anxiety and motor function using the Unified Parkinson Disease Rating Scale – a too
What if... (Score:2)
she does have a very short stub (Score:2)