It creates the incentive that I create controversial content to receive more clicks. The by-product of that is that when I really think something I add a preamble: “authentically, genuinely, not playing a devil’s advocate”
How to protect myself?
I would like to use my personal AI assistant, working title “Reality Filter” that will live a Chrome extension or some scripts on my computer / mobile phone and process the bullshit and feed me the distilled content that is optimised for my health / wealth / happiness / wellbeing rather than shareholders profits, advertising budget, time on the site.
I agree with Linus Torvalds “talk is cheap show me the code” but maybe there are some existing projects in the space?
Welcome to HTC @marsrobertson. In terms of consuming content optimized for you, you might have a look at using a good RSS feed aggregator.
RSS standard and its use is de-emphasized by the big players, as it breaks open their walled garden, but some of the best sources on the web and most personal blogs etc. still have RSS feeds to subscribe to.
I’m not aware of any projects that filter content in a trustworthy way like you have described. However, you might enjoy reading Cal Newport’s “Digital Minimalism”. The book recommends a few software tools/tweaks to conserve your time/attention, and presents strategies to help you find relevant and high-quality online content much more efficiently.
One recommended change is to batch deliver notifications so that your attention is not frequently pulled to your smartphone, interrupting your focus throughout the day. Instead of consuming more news, more social media, more screen time, and trying to find or build tools to make this process more efficient you can preserve your attention span and search only for specific information you are interested in when you have a natural break in your day.
Another recommendation is to pay for a news subscription. News agencies that do not charge directly for their services instead load their websites with trackers & ads, they will write sensationalist headlines to grab your attention, and are incentivized to waste your time by keeping you on their website as long as possible. I recommend not reading most national and international news and instead consuming most of your news through a reputable local news agency. Local news is almost always far more relevant to your life, is not filter bubbled, has greater journalistic integrity, and local editors are much easier to contact.
I subscribe to my local news paper and use pocket on Firefox to save articles I want to read throughout the week. Then, every Friday, I sit down and go through the news and articles for the week in only an hour or so. Instead of many hours spent throughout the week browsing for news and searching for more information, I catch up on the news after all the “breaking” stories have been properly sourced and fact-checked saving lots of time and outrage. Few news stories are so relevant they demand your immediate attention, and if there is one you’ll hear about it because it interrupts the normal course of your day.
Breaking news finds me anyway from https://news.ycombinator.com/best - some really important non-tech news lands on the homepage anyway.
I still believe there is potential for a tool that will be designed for the ground up for happiness, health, wellbeing as opposed to maximising time on the side.
Tangential rant: I work as a developer. A new Ubuntu virtual machine, opening Firefox: bombarded with news. then RDP to EC2 on AWS, opening Edge browser: bombarded with news. My head is literally WTF they were thinking, these are tools to work, not to get a shit show information when opening a new tab. These defaults and business models are ridiculous.
I think it is a proper good idea, but I haven’t heard of any projects that facilitate this though probably a whole bunch come close is some way or other (maybe not AI-based). Worthy of an investigation.
Netflix awarded a $1 million prize to a developer team in 2009 for an algorithm that increased the accuracy of the company’s recommendation engine by 10 percent. But it doesn’t use the million-dollar code, and has no plans to implement it in the future, Netflix announced on its blog Friday.
But for Facebook. Is Facebook likely to fund it?
Or maybe they already run internal competitions and train the AI to select relevant items?
Taking it a step further…
My personal AI detects my mood. Depending on my mental state, I’m presented with matching content.
Just like a restaurant that takes my blood sample and AI automagically orders the food for me, with the right proportions of each vitamin to put me back into the optimal state.
You might enjoy reading User Friendly by Cliff Kuang. There are several chapters that talk about tech that comes close to what you describe, specifically Disney Magic Bands, and experiments by Carnival cruise lines to tailor customer experiences. I don’t remember all of the specifics of the examples given in the book, but they talk about how the physicallyDisney Magic Band project is far more comprehensive than its current implementation. Disney can use the Magic Band to track customers’ exact location through their parks and use high quality surveillance cameras to record the customer’s experience for them. They can use AI to select the best photos of the customer and their family during their time at Disney and have those photos & videos bundled and ready to sell to the customer on their way out of the park.
The book also talks about using audio recording and AI to infer the customer’s current mood and desires in order to tailor Disney’s customer service scripts and to target opportunities for upselling.
“You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.”
Product idea
BUT for digital content that lives on my device
It can live as Chrome Extension, it can live as background process, Jarvis, Siri, Alexa, Cortana, Google Assistant, you name it… The most significant difference: it is working for me, not for them.
Success metric = lifetimes per day
It costs $3000 to save life, malaria has the best ROI measured in lives per dollar:
Can we do better?
Back of a napkin math, basic assumptions:
80 years lifespan
3 billions daily users of Apple / Google / Microsoft / tech in general
5 minutes saved per day
24 * 60 * 60 = 86,400seconds per day 365 * 86400 = 31,536,000seconds per year 80 * 31,536,000 = 2,522,880,000seconds per lifetime 5 * 60 * 3,000,000,000 (3b people) = 900,000,000,000 (900b)seconds saved each day
That is 356.7 lifetimes every day.
5 minutes is a conservative estimate, the “autoplay” on TikTok on YouTube can consume hours.
I think that US Congress will be too slow to regulate FB / Twitter / other media platforms.
I think that the innovation has to come bottom-up.
I think that people should own their devices and be able to run any software their desire (not always the case)
I think it is within our technical ability to create PERSONAL REALITY FILTER - ability to optimise on the client side what type of content is serving my health, wealth, wellbeing.
I think that if this problemis presented to Apple / Google / Microsoft they will find every excuse in the book to say “impossible, against fair competition, etc”
I still believe it is possible
A piece of code that is working for me, not for them.
Facebook newsfeed, Twitter newsfeed - they own the data - they run their AI to profit the shareholders (maximise engagement, time on site, media impressions).
My personal AI could do it better - with totally different incentive structure.
I think it is possible.
If you can explain something in 2 sentences then hell yeah for sure it is possible.
Is anyone working on it?
Next-level would be real-time BCI that measures my mood and depending on my current mental state shows me the stuff that I should see.
(just like taking a blood test before going to restaurant and AI picking up a meal to properly balance my nutritional content)
Can recommend checking out the Fediverse, which is just such bottom-up. It is social networking, which is much more than media. It’s vision - at least as I advocate it - is to become a Peopleverse (which has entirely nothing to do with the lifeless vision of a Metaverse).
I don’t know Fediverse projects that work with AI (as yet), though you might find this interesting:
Free and open-source tools often lack a business model. Maybe crowdfunding, maybe philanthropy, maybe social impact but for-profit corporations are way more efficient
(debatable subject, maybe big tech is already working on it and A / B testing happiness, health, wellbeing, love metrics)
Don’t think it is too hippie. The ways in which we design our supportive technology rubs off on the people it serves. By taking positive psychology into account in our designs we can at minimum ensure that we don’t overly restrict the richness of social interactions we have online in comparison to how we have them IRL, create safe spaces, but we can go further and encourage people to bring out the best in themselves.
The Wikipedia page is not the best. I co-maintain a couple of ‘delightful’ lists relating to the Fediverse, like:
In terms of adoption it is obviously nowhere near the big social media platforms. But there are millions of fedizens, equating to the population of a decent-size country. Which is not bad. It is all free software, and that can have poor UX/UI, yes. Some apps are well-productized like Mastodon (Twitter alternative), Pixelfed (Instagram alternative), Peertube (Youtube alternative) and Lemmy (Reddit alternative). I like the UI of Mastodon much better than Twitter with all the garbage content the algorithm throws at me. I am in full control of my timelines, which are chronological and only show stuff of people I chose to follow and their friends. Here is a screenshot:
Most special are the unique culture and values. In many ways the Fediverse has the vibrancy of ‘the early web’, and I consider it a field lab and playground for humane technology. It doesn’t come without its challenges, though. There are no corporations steering its development with a ton of money and resources. It is all grassroots and like with free software a “Tragedy of the Commons” is something to be very wary of.
Free software often start as hobby projects of a developer. Then, with success, it is hard to productize them, let alone get a sustainable income from them. But there are shifts in the landscape. There’s also a lot of positive developments. Most of those you aren’t aware of, if you aren’t exposed to it.
It is ironic. There’s the saying “software has eaten the world”, but more accurate nowadays is “open source software has eaten the world”, and with recent debables like log4j security issues many corporations start to realize that they depend on OSS and also should support it more. In the EU at all levels of government there’s some big policy shifts gaining steam where a (F)OSS-first approach is mandatory. This will create many incentives and opportunities for more sustainable businesses to occupy these new niches that open up.
And besides that there’s increasing awareness that when we let Big Tech have their way we’ll move to absolute shithole dystopic sci-fi future
I really wish I was better at executing on all of these ideas.
Currently my full focus is on basex.com - new definition of value - accounting for negative externalities and positive cobenefits of large corporations. Related to that - new metrics for carbon credits markets. New values, metrics and new incentives of social media - I strongly resonate with this vision but I do not have spare mental capacity to engage fully.
I don’t know. Have been twitter user for a bit, but have never witnessed a substantive discussion. Mostly a good tweet or tweet chain is followed up by a ton of low-value responses. And also I find the number of interactions / engagement (if you aren’t an influencer probably) to be very low. You should try the fediverse and be astounded how different that is, with only a fraction of the people-base.
WTF they were thinking, these are tools to work, not to get a shit show information when opening a new tab. These defaults and business models are ridiculous.
