Ledger of Harms
ledger.humanetech.com
ledger.humanetech.com
Now the Amazon folks came along with the "Alexa prize". Whoever creates a bot, that people want to spend the longest duration of time with, wins. So maximal user engagement is seen as measure of intelligence.
I find it really weird that our tools seem to seek to waste the largest amount of time for us. I would've thought, that the ideal assistant would minimize engagement time, while maximizing utility; and it also seemed to me like a product like that would easily blow the time-wasting competition out of the water.
But it doesn't seem so. Maybe too early in the development...
Yeah and it might seem fun initially, yet in the long term shit like Alexa and striving for artificial ‘human-like interaction’ just creates more alienation.
...and alienation and isolation, and an atomized society, only benefits capital. Our current mode of production aims to keep workers easily replaceable, with projects siloed and in rigid knowledge hierarchies that make workers dependent (and keeps the propertied class propertied), instead of enabling and fostering the individual and collective intellectual and emotional growth and skills of humanity.
Instead of AI to keep us ‘company’, we need systems that enable humans to have maximum transparency, agency and participation [1].
Platform cooperativism and ‘protocol cooperativism’ (through projects like http://valueflo.ws and holochain) are imo the most important shifts of our time.
[1] https://www.theguardian.com/environment/earth-insight/2014/j...
Yeah, I think that we already have maximizing for user engagement in social networks. That didn't turn out that great so far, at least in my opinion. Now we'll have have bots actively spreading most polarizing, rage inducing topics so that engagement is high.
A longer Alexa sentence != longer engagement time, which I don’t think they’ve figured out yet.
They track the rate of negative feedback and high negative feedback will kill it over time.
Where do they state this? It seems like maximal user engagement = maximal profits and that's all Amazon cares about.
https://www.fastcompany.com/90590042/turing-test-obsolete-ai...
It was infuriating as a teenager just trying to get the job done so I could slack off with the cute crewmembers or read the backs of the video game boxes for the umpteenth time, but now that my social circle has dwindled to wife, children, acquaintances, and work I get it. A deep creeping loneliness when you have no one to talk to and in my case all the money in the world that younger me never could have imagined.
It doesn't shock me at all that Amazon would want a way to engage older, generally more wealthy, people and I am sure their expectation is that if successful it would lead to significant sales increases or at least mind share capture for their platform over their competitors. Here's to hoping that they are successful for the sake of the lonely in the world and that they enjoy their 1000% increase in diet coke sales.
Or perhaps I'm overestimating what people know about current 'AI' systems. The less you know about or understand a subject, the more mystical and sometimes interesting it appears.
Customer support is a cost center for most companies, so I can see where the strategy of a) minimize costs and b) provide just enough support that the FTC doesn't get involved while making the experience frustrating enough (including the cancellation experience) that most users accept crappy service.
I fcking hate* those "voice assistants" that companies use on their "support" lines, to avoid hiring actual people. Almost everyone would rather punch "1" or "2" to avoid them.
And I doubt that anyone calls them "just to chat" although maybe if I listened to the call logs, I'd be shocked.
Yes and:
I'm mostly against voice tech. Just feels weird. Still frustrating to use, after 30+ years.
My skepticism that voice would ever become a normal thing was mooted by observing my son play a Living Books CD-ROM adaptation of a Mercer Meyer's Little Critter book. https://archive.org/details/littlecrittergames
Critter asked a question. My son answered. Then nothing happened. My son was hurt, disappointed, confused. No different than if a human child ignored him.
My totally obvious prediction:
Voice UI will become normal for the generation that grows up with a Critter that understands verbal responses.
And to your point, those Alexa style agents will remain their life long companions (pacifiers) into old age. Just like my 84 yo mom's relationship with the TV.
"Just get the truck done" could probably have been our motto.
I'm assuming not, both because hard, missing corpus of training data for high quality q&a understanding, and minimal demand. But maybe I'd be surprised!
Edit: Should have included a link: https://www.home-assistant.io/
It works well with HomeAssistant to change the lights, etc.
and
Voice2json - http://voice2json.org/
I've been playing with lately. Mycroft is pretty solid out of the box (or at least, promising)
The idea (absolutely correct, IMO) was that the shorter the visit, the more successful the search was. I.e., if they provide exactly the desired result at the top of the list, the user would promptly click off to their desired destination, whereas if they gave poor results, there would be many more search actions.
However, they have long since strayed away from such sound reasoning.
One that stuck out to me is "Technology integrates and often amplifies racism, sexism, ableism and homophobia". This seems unreasonable - all of these things were substantially worse before technology got involved. Technology seems to blunting the sting, enabling more effective organisation among minorities and promoting a better understanding of the problem.
It seems likely that many of these problems were much worse before the big social media companies too.
The problem is not the technological concept itself but where and how we consume it (e.g every 30 seconds). Keep in mind Facebook started when mobiles were somewhat limited and children had flip phones if at all. And on those, “background apps” were scarcely a thing.
Nowadays, people are like zombies—and more time is spend wasting away in front of multimédia than providing value to the world.
The "good" doesn't have to be significant for us to become addicted. There are some upsides to smoking cigarettes and other self-destructive behaviors, I'm sure.
Replace that spot where you go to smoke with a foursquare court and dare your boss to tell you you can't take as many foursquare breaks as the smokers get. Foursquare is fun and you and your group will invent your own rules and it'll turn into your own game that's way more fun than smoking cigarettes. Most places now, you can smoke a joint or a weed pen if you really need the physical act.
There's this idea that if you smoke you get extra breaks - bring it up in a reasonable way with any manager and you can have those breaks, too. And thank the few folks sacrificing themselves so that the rest of us can say we get as many breaks as they do.
Edit: a benefit I miss is excusing myself from an uncomfortable situation because I "need" a cigarette. But, without that crutch, I developed better GTFO strategies.
I'm not really sure that they were. Between the late 60s and the early 2000s they seemed to go way down in western society. In the 2010's they started to pick up again with the rise of smartphones and social media.
I have always operated under the assumption that the amount of (latent) allophobia in a given population is pretty much constant (or rather, a sawtooth wave around a hardly-changing trendline). The reason that we don't have violent clashes every week is because society is big enough that the oppositional groups don't come into direct contact that often. It seems that social media facilitates that direct contact (and algorithms actually drive it, because engagement), so we see more of the ugliness. But I'm not convinced that more overt allophobia automatically implies that there's more allophobia in total, just that it's more readily expressed.
But I realize that this is all conjecture, because there is no way to accurately measure latent allophobia anyway. It's just my personal interpretation of complex societies.
I'd agree -- the OP ledger, despite seeming well-intentioned, seems completely imbued with the current on-campus value system, wherein something isn't defined as really bad unless it harms BIPOC and LBGTQIA people more than it harms others.
If we define the arrival of the internet as roughly the early/mid 1990's, America as a country has gotten extraordinarily less racist and homophobic since then. For example, 33% of white people in the South approved of interracial marriage in 1991; today that figure stands at 93%[0]. Similarly, gay marriage was completely eschewed by political leaders, including Democrats, until the early 2010s (e.g., Obama was firmly against gay marriage, at least on paper, when he first ran for President).
None of the above is to say that we now live in a colorblind or non-homophobic society, I just find it odd that they're claiming that tech is making people more bigoted, when the opposite has happened since the Internet has been a thing -- but this is somewhat understandable given that their target audience has been conditioned to see stamping out bigotry as humanity's highest calling.
[0]: https://www.forbes.com/sites/tommybeer/2021/09/10/record-hig...
"1. An inaccurate statement or statistic believed to be true because of broad repetition, especially if cited in the media.
2. (originally Canada, US) An interesting item of trivia; a minor fact."
I kinda looks like a fact but isn’t.
They seem to recommend just the right thing to “okay lets watch this one last video” wasting hours of my day.
Though, I did have to manually ignore a bunch of clips-from-tv -shows type channels.
- turn off Youtube history, there will be no more suggestions based on your recent viewing activity to keep you engaged. Also, searching for a sink review will not result in Youtube suggesting sink-related videos for the rest of eternity.
- install browser extensions: Clickbait Remover for Youtube (sanitizes video previews to remove clickbaity look from them) and DF Youtube (allows you to disable comments, related and other sections on the page, leaving only the video itself to focus on).
Now you should be able to deal with this addiction in no time.
My main (practically only) use for it is to watch amateur-produced special interest shows that are done on a regular basis. Usually it seems to involve automotive bodywork or subsurface mining with the occasional computer show.
The channels that move to having sponsors or getting a lot of free stuff are for the high jump (although it's absolutely a riot to watch Heather Heying shill for some sofa company, hating life all the while). I think the best ones are done by pure hobbyists.
Otherwise my current plan is to build a library of worthwhile videos via youtube-dl and DVD transfers. My guess is that the ad environment and much heavier copyright policing on youtube will make much of it unwatchable over time.
People who profit from encouraging poor dopamine hygiene will have an outsized influence on the people they've hooked, doesn't matter if you're managing an ad campaign in the US or part of the opium pipeline in Afghanistan.
I originally clicked the link hoping that a "ledger of harms" would chronicle the worst offender for each harm. Not inciting anybody to anything, just making it known that if someone had some adversarial behavior in mind, then it would be best directed at such and such an organization. That's what I would call violent agreement.
It basically blocks distracting apps by default and gives you a barrier to open them like shaking your phone for a couple of seconds! Eager to hear what any of you think of it!
So, a special kind of list indeed.
Of course I can't (and wouldn't) force people at gunpoint to log our of their social media. Nobody is suggesting that.
The article ("ledger") attempts to make an argument for a societal change. How would one start changing society, if one is not allowed to start gathering arguments/evidence in favor of the change one wants to see?
"People be people", sure; if there wasn't a trillion-dollar industry profiting from people being hooked on social media and applying every trick at their disposal to keep people hooked, then you would have a point. No, it's not in people's nature to be hooked on scrolling Instagram. It's an entirely artificial problem that we have created together.
> Children who have been cyberbullied are 3x more likely to contemplate suicide compared to their peers.
> Children who experienced cyberbullying during their adolescence were significantly more likely to engage in risk-taking health behavior as adults.
Yes, bullying is bad. Now kids can get bullied online. Such is the way of things.
> About 40% of 13-17 year olds reported it was “normal for people my age to share nudes with each other.”
.. so? People used polaroids before. It was harder so it was done less often, but done it was.
> Children who see videos of child influencers holding unhealthy foods consume significantly more calories than those who see influencers holding other types of objects, as clearly shown by experiments using randomized control trials.
In other news: researchers rediscover priming is a real effect.
> More than half US middle-schoolers cannot distinguish advertising from real news, or fact from fiction.
That's how its always been. Don't most people have a memory of being a kid around that age, and have your parents blow your mind by suggesting that not everything you see on TV is true?
> 77% of teenagers get their news from social media, with 39% stating that they “often” get news from celebrities, influencers, or personalities, according to a survey of over 800 teens aged 13-18.
Ah, so they should get their news from the TV or newspaper instead. Much more reliable, right? All news is dogshit, and various levels of lies. If you've ever been the center of a news story, you know that they literally lie and pull stuff out of nowhere constantly. Its made my jaw drop before.
> 30% teens report that they pay “very little attention” to considering the source from which they are getting their news on social media.
So 70% pay more than "very little attention" (phrasing!). I'd say most teens are actually better at discerning scams/obviously misleading content online than non-techie 50+ year olds, as long as it doesn't align with their biases (but that's a universal problem).
> Several self-harming videos have been circulating on TikTok, from the "Skull breaker" challenge to the "Cha Cha Slide" challenge (which involves repeatedly swerving a car across a road in time to music).
BREAKING NEWS: what do the bracelets around your child's wrist mean? Could they be engaging in secret parties where they take drugs and suck dick?
This is fearmongering.
> The number of US teenagers who are online continuously is increasing at a dramatic pace, almost doubling from 2015 to 2018: 24% to 45%.
Because the internet is taking over the world.
You got a big list of arguments for one viewpoint. You might want to point to evidence of your opposing view instead of dismissing the science without addressing any of its claims.
I’m strongly in agreement with the goal and the sentiment of the project, but not with its execution. We need a lot more epistemic humility and contextualization for this to not just seem like just another political project.
This is just first-grade populistic crap. Yes, there's various levels of lies and shit, but that doesn't mean they're all equally bad. Being able to discern the quality and trustworthiness of a source is an important skill, and it doesn't seem you are willing to master it.
This is really the only way to even get close to the truth. Every time I have read a news article on a subject I am an expert on, or about me, or about my family or friends, it was so glaringly wrong that anyone else who read it is now less informed than before. Sometimes not just in the facts, which might be mostly correct, but in the false narratives they present.
> > About 40% of 13-17 year olds reported it was “normal for people my age to share nudes with each other.”
>.. so? People used polaroids before. It was harder so it was done less often, but done it was.
You state that social media makes it easier, but that is exactly the point of the OP article. If it is harmful behavior, and social media makes it easier, then social media is contributing to harming kids. Right?
If you want to shrug it off, at least state your argument honestly, like “I don’t think it’s a problem that 14–year-olds share nudes between them, so it’s not a problem that social media makes it easier”.