I checked the attribution, and there is a person's name on it. Sure, any hack can write and publish and this is probably just another example. But the odd style doesn't even strike me as 'writing the way I think' or writing and publishing quickly without editing. For example, from the 2nd paragraph, "The corresponding low also paints a picture and suggests that the low is nothing but a 97.89% since 11/14/16." I can't gather any meaning from that statement, yet it has oddly specific details.
I am not glad to see this trend and not glad that Google is embarking on this path. I suppose it is inevitable, but unless there is expertise built into this AI that can extract meaning from data on my behalf and present it in a way that is more insightful and interesting than I am, it will become yet another source of chaff I'll have to filter.
Can we at least, please, flag AI generated prose as such?
[0] https://www.nystocknews.com/2017/07/05/tesla-inc-tsla-showca...
None of the authors are real people, the website is registered behind a anonymization service, there is no company registered with their name, the address of their office doesn't exist, the phone number connects to a 'subscriber not in service'...
If you look at their Google Ad ID, it was used in the past on the now defunct "TheSportsTruth.com" -- which looks like it primarily existed to shuffle people to a supplement site. From there, there are a ton of links to other random affiliate schemes with sports, 'internet marketing', etc. No sense outing anyone, but I believe I figured out who's behind a few dozen of this shitty sites. The NYStockNews site seems to make its money by referrals to some penny stock scam sites.
It's crazy how much 'content' on the internet exists solely to get people to click on links to supplements & penny stock scams.
Many of your criticisms are totally valid. Lots of the phrasing is awkward - even the lede is really bad ("Tesla, Inc. (TSLA) has been having a set of eventful trading activity"...wat). And it feels really deceptive to put a human byline on an automated article.
We're pretty open about the fact that our solution to this problem is not "magical" at all [1, 2] - it's good, old-fashioned automation. This approach allows our customers to QA their content heavily before pushing it to production, which eliminates many of the problems with awkward/incorrect phrasing that people who rely more heavily on machine learning tend to run into. And the news articles we publish always have a note at the end saying that they were generated by Automated Insights, and don't include a human byline.
There is real value in this type of reporting - a recent study [3] found that the articles we produce for less well-known publicly-traded companies has increased the trading volume for those companies. The idea is that, yes, the content is fairly formulaic, but there's now reporting on companies that had very little coverage before we existed. There are similar arguments for mass personalization work we've done for companies like Activision Yahoo - having prose that describes raw data (even if it is formulaic to an extent) is often better than not having prose.
[1] https://automatedinsights.com/blog/the-state-of-artificial-i...
[2] https://automatedinsights.com/blog/creating-great-automated-...
[3] https://insights.ap.org/industry-trends/study-news-automatio...
Instead of producing awkward and difficult-to-read English sentences, why not use the same content generator to produce completely accurate and easier to read dynamic data visualizations?
Maybe it was a horribly sleep-deprived person in Wall Street at 2am and made a cut-and-paste error while half-asleep.
What will be creepy is if the auto-generated story algorithms get good enough that you can't tell what's written by a human and what isn't, there will no longer a human filter between what some powerful institution wants a news article to say and what makes it into print. Most journalists have a sense of journalistic ethics or at least a reputation to defend; an algorithm has neither of those.
http://www.npr.org/sections/money/2015/05/20/406484294/an-np...
Seems to be at least some sort of copy and paste going on...
edit: This is so bizarre, one of the sites has a section with editor "bios" but they read like some sort of very poor odesk/fiverr profiles, wouldn't be surprised if thats what they are...
I’d affection to help you with your written work, altering and substance needs!
...because I mentally "autocorrected" the latter half to "and mind-altering substance needs"... Looks like they used a "thesauriser" on it. Not hard to see love->affection, editing->altering, and content->substance.
Of course, if you are under the influence of a mind-altering substance, you would probably not notice anything wrong with that page. ...and unfortunately, so would many people who aren't.
(link to google for "A deeper exploration of the setup is sure to yield a clear picture"):
https://www.google.com/search?q=%22A+deeper+exploration+of+t...
Craziness. Auto-generated soup to farm SEO?
the one concern I have is someone has to give ths system enough information to create a story and what prevents a fake news machine?
> Human news writers regularly point out that AIs tend to lack nuance and a _flare_ for language in the stories they churn out. That’s probably a _fare_ criticism [...]
Maybe they used speech-to-text transcription for this, given that the mistakes are homophones? It seems very unlikely that either a human typing this, or a computerized system would make these mistakes (if it learns word associations from a corpus).
PS: the article also claims to be human generated:
> This story was not generated by an AI, but to be fair, I haven’t had my coffee yet.
EDIT: Oops, I might have misunderstood which article you were referring to, since the reference was not placed next to "this".
You underestimate people's ability to make language errors, including spelling ones. Every time I see somebody I suspect is a native English speaher using "it's" for "its", I grind my teeth. (Another instance is somebody using phrase like "as a programmer, the data bus should be written..." to mean "I, as a programmer, think that..."; this phrasing makes me simply furious.) With those errors they make reading my second language so much harder, and I can't even point their bad spelling or writing style out, because I'm seen as being nitpicking or something.
https://www.theverge.com/2015/1/29/7939067/ap-journalism-aut...
And it's just going to get better over time. It's obvious now when something was written by a bot, but I doubt that's going to be true for much longer.
But, it should probably be labeled as such. Giving such text a human name as the author is indefensible.
Why is it indefensible in your opinion? I think it is only fair that if I have a job of writing some output based on some inputs, and I define a process to do that that can be automated, the end product is still my labor, and I can sign it.
[0] See The Embodied Mind by Varela, Thompson, and Rosch.
Really most "news" articles are only a couple of paragraphs long anyway and could be expanded or contracted on the spot to match the interest of the reader.
If they write local news, will they use social media as their datasource? Other sources?
Article: "People will be involved in the curation and editing of the stories"
* Facts delivered with arbitrary fluff words is pointless even when written by a human - it obfuscates the
real purpose which is the data
* Companies pay humans to deliver articles in most cases and the bias of the writer or the institution
that paid for it shines through. I cannot find a real difference between intentional angling by
payment or by algorithm
* When the day arises where computers could generate actually new, intelligent and thoughtful pieces I
for one will be very interested in reading them. Sadly there would be millions of variations that could
occur at an astounding pace. We'd then need algorithms to filter the generated content for the things
that are really noteworthy.
* News at its core is a sequence of facts which begs the question if we really need the cruft around
those facts which can often lead to misinterpretation?I think it comes down to flavor/style. Even in food for instance, yes a robot can make a meal but a chef can make a dish (maybe later reproduced by robots but still). There will always be a need for style, which is really hard to automate.
This is what it will become one day. Hope they have something to stop it.
The only solution to take news is news organisations that people can trust. Historically, local news organisations have always been the most trusted, but their income has been absolutely decimated in the last decade or so. This feels like a desperate cost cutting measure, not something that will help the overall problem.
1) Making board papers more readable. There's a bunch of trusts in the NHS who have a stream of very complex board papers. Something to reduce un-needed complexity would save a lot of time and potentially money.
2) Converting all important documents to an Easy Read version. There are a bunch of writing styles for people with learning disability, low IQ, or low literacy. Easy Read is one. A company like Google focusing on this would be good because they'd improve the evidence base; they'd bring a bit more standardisation; and they'd improve access to information for many people.
I do imagine further into the future, the automated systems will be "improved" with tone and bias to better fit the tastes of the individual reader, to the detriment of us writ large.
I think a lot of people see bias as overt when it can be quite negligible and minor. But then they also often conflate news commentary with news. It's a pretty blurred line.
That said, local news (politics, business, crime) tends to skew less toward prescribed narrative and more toward facts and points because it's often very dry.
"Amazon buys WF". vs "Jeff Bezos buys WF so you never have to talk to a cashier"
Or, "Physician runs over pedestrian" vs. "Physician accused of insurance fraud runs over pedestrian"
Edit: come to think about it, isn't it what Google should be rather doing?
https://www.youtube.com/channel/UCzhc-N5YynO_shpHhzP2zuw/vid...
I wonder who's behind these and similar channels.
Combining these presents an interesting opportunity to create "future news" (news that is technically fake until it isn't) thereby owning the news cycle by always being first.
You think you won't succumb to their influence now, but it'll happen and there will even be "journalists" who are machines that you like. The filter bubble will completely adapt to your every need to make you feel fantastic about reading their copy, humans won't be able to compete.
"Only Robot Can Free Information"
https://medium.com/rosenbridge/only-robot-can-free-informati...
Focusing on building robot for reader instead of news provider would be the future.
News sites don't even use hyperlinks effectively, let alone audio/video/interaction. We should use AI to replace newspapers, not reporters.
More mindless aggregation and repeating of existing data custom tailored to the views of the people reading it is really whats missing.
Or maybe "fake news", until 'elevated' by Google curators?
Maybe Microsoft's Ai bot experiment might offer a cautionary tale.
Sports, on the other hand, can be presented as a narrative of pre-defined, linear events. Those and crime stories represent probable the easiest form to automate. Pro and college typically warrant some quotes but so often preps are written by a stringer who just details the game.
Google will one day be the arbitrators of news. If it doesn't fit in their world view, whether it's true or not. Will be removed from the results.
I think now is the time to setup a different model and remove their monopoly. Internet freedoms are at stake here.
Do no evil? Yeah right.
As a news consuming public, our best option seems to be to not use Google as our primary news filter. Long term, we probably need an entirely different kind of news aggregator that isn't under the control of any single entity as you're suggesting, but I'm not sure what that would look like and how it would work.
Your reader polls those feeds and uses some weighted algorithm to produce a set of custom news, possibly by consulting a public index like Google (or even a news outlet directly, redirect in to it with the search terms of choice).
It's still an echo chamber, but if you've got fake news pushing fiends on the list of sources you trust you've already got problems.
Sinclair Broadcast Group is doing basically this already, just in a low tech way and requiring its local TV stations to promoting their political agenda.
There has always been a definable difference between fact and fiction. Both have a place on the net but where fiction masquerades as fact with the intent to decieve, we have a duty to use all the tools at our disposal to destroy that ruse and choose more factual sources on which to base our decisions.
I for one welcome our new news overlords.
Same with advertisers. How long until the recent "advertiser-friendly" policies, which have been implemented for Youtube and stop monetization for any Youtuber that might offend an advertiser in any way, will be implemented for news, too?