The fundamental flaws in modern social media infrastructure are nothing new, and nor is the idea of using blockchain and other decentralized technologies to help remedy them. But never before have these issues risen to the fore as they are currently. Never before have so many people been so aware of the ethical and practical weaknesses at the heart of Big Tech’s systems for controlling our communications.
The American people’s collective jaws are still somewhat agape at the speed and oomph with which Apple, Google, Twitter, Facebook, Amazon and others have recently acted to combat online discussion perceived as encouraging right-wing anti-government violence. High-profile actions like Twitter suspending U.S. President Trump’s account, and Apple and Google and Amazon concurrently deplatforming Parler got everyone’s attention. Just as disturbing was Google threatening to remove Facebook alternative minds.com from the Android Play Store if it didn’t remove content Google found offensive.
Minds managed to stay on the Play Store by hastily gutting parts of its site like the commenting functionality. And it’s definitely not an alt-right, MAGA-oriented social network. Bill Ottman, who founded minds.com, is pretty clearly left-wing in orientation, though he keeps his personal politics out of his company’s operations. Along with a lot of fantastic content, Minds.com does have some nasty commentary on it, and more QAnon bulls**t than I would ever want to read. But, just as on Facebook, I don’t need to go to those pages except when struck by a particular bout of perverse curiosity.
And the internet, of course, also has a lot of nasty stuff and a lot of QAnon BS on it, which one can readily find using Google’s search engine. Why is it that Google considers it ethically okay to point searchers to QAnon pages, but also feels an “ethical” need to screw with minds.com because it has a minority population of QAnon members?
Who’s going to solve social?
One can sympathize to some degree with the folks who run these Big Tech companies. They’re in a classic “damned if you do, damned if you don’t” situation, which none of them foresaw before they got so huge.
If they allow offensive content or folks arguing for violence in a not-quite-illegal-but-pretty-nasty way, then they’ll get blowback when it turns out content they hosted played a role in some actual horrible acts.
If they start banning content, inevitably they’re going to ban stuff that isn’t actually extraordinarily dangerous, and suddenly they’re evil opponents of free speech.
And with the volume of situations and people they’re dealing with, how can these companies consistently figure out the right place to draw the line, especially when society as a whole doesn’t have a consistent or coherent idea of where to draw the line.
Twitter, generally the most forward-thinking of the Big Tech titans, has appointed a small team to develop a decentralized social media protocol. The idea, apparently, is to make Twitter one among multiple services using a common standard protocol, and to move controversial issues like content moderation and banning from the Twitter layer down to the decentralized protocol layer. The aspiration is admirable but at the present time it’s not at all clear how the fundamentally centralized logic of the Twitter business model is going to harmonize with the radical degree of decentralization that would be needed.
My own strong feeling is: Twitter is not going to solve this, as good as its intentions may be. The centralized Big Tech world is not going to solve this. Government is not going to solve this either. It’s far too slow-paced and obtuse to deal with fast-evolving issues that are highly complex in both human and technological dimensions, in a way that’s compatible with the subtleties of Western constitutional and business law.
The mainland Chinese government has “solved” the issue of Big Tech and social media in its own way: by exerting explicit and detailed government control. This works fine in the modern Chinese context, but only because this context doesn’t involve Western-style expectations of free speech or business autonomy. If you're a mainland Chinese tech CEO, you don’t need to do a lot of soul-searching about what’s ethical or not. You just have to accept government guidance. Elsewise, at best you’ll end up like Jack Ma, laying low for a while and then shifting your focus to philanthropy.
The good news is: The elements of a truly workable solution to Western society’s social media problem are reasonably well known and not incredibly mysterious.
The more complex news is this: The blockchain community knows how to fix these problems, but it is not remotely as well-resourced as Big Tech, is not exactly a darling of the government and has a better track record making cool tech for elite geeks than apps that give warm fuzzies to the average TikTok user.
Golden opportunity for blockchain
We have recently seen another example of an acute social problem where the blockchain community held many of the keys: the COVID-19 pandemic. Blockchain technology is tailor-made to provide secure, anonymized tracking and tracing of COVID-19 infection.
Technologically, there’s no good reason every Western citizen with a smartphone didn’t have a blockchain-based COVID-19 tracking and tracing app on their phone. These apps could have analyzed personal medical data using secure decentralized AI and fed everyone’s data into agent-based epidemiological simulation models running on decentralized platforms.
Had this happened, the U.S., Western Europe and Latin America would have likely squashed COVID-19 as rapidly as the Asian countries (where use of less privacy-preserving track-and-trace apps tended to be socially accepted). Had blockchain played a key role in helping the West dodge COVID-19 while preserving privacy, this would have been a pivotal moment for decentralized technology. It would have proved in a wonderful way that it’s good for something besides geeky financial gambling and money laundering.
The obstacles to using blockchain to help with COVID-19 weren’t mainly technical. In fact, during 2020, an interesting variety of software tools were created using secure decentralized technologies to do track-and-trace, personal health evaluation, epidemiological modeling and a lot more. But the adoption of such software has been minimal so far. This was a golden opportunity lost.
Decentralized social media is an even larger and more critical opportunity. As tragic as the COVID-19 pandemic has been, it’s soon going to be squelched by vaccines. There is no vaccine for badly designed social media networks on the horizon. The negative impact of Big Tech-controlled social media on humanity has far greater destructive potential than the coronavirus pandemic.
There’s a decent argument that Big Tech social media, by fanning flames of hate and tribalism, could significantly increase the odds of the annihilation of the human race.
Academic boffins like Nick Bostrom are worried about superhuman AGI repurposing human molecules to add extra memory capacity to their hard drives. But there is far more evident risk in Big Tech’s use of narrow-minded machine learning algorithms to maximize shareholder value at the expense of human well-being, intelligence and compassion.
Components to save social media
The basic principles needed to create better social networks are not too hard to see:
- Open-source code: Code should be open source so that everyone has the ability to inspect and understand the algorithms that are putting images and videos in front of their eyes, people into their friendship circles, and texts and ideas into their minds.
- Decentralized control: Ownership and control should be decentralized, so no company and no government is tasked and burdened with calling all the shots.
- Democratic decision: Major decisions about social network structure and dynamics should be made democratically. And if democratic decision making doesn’t converge on a result, it’s ok if a social network splinters into multiple ones, just as Linux, Bitcoin and Ethereum have forked into multiple versions.
- Explainable AI: Recommendations of connections and content should be made using open-source AI that is created via democratically governed projects, and that has the ability to explain the reasons behind its judgments. These explanations need to be shown routinely to users in a way they can understand.
The technologies to manifest these principles are here right now, albeit at varying levels of maturity.
In SingularityNET, we have been working on blockchain-based AI tools suitable for serving as the cognitive core of decentralized social media networks. Your social media and Internet usage data should be owned and controlled by you, and the AI algorithms used to model this data should be under your control. You should be able to inspect and interrogate these models and understand what they are thinking about you and why they are recommending certain things to you and not others.
Transparent, explainable AI applied to a person’s social media and Internet usage data has the potential to be a powerful tool for self-understanding and self-growth. An AI watching what I do will very likely understand some aspects of myself better than I do, and this is of interest to me because I would like to better understand myself.
It’s easy to imagine a smarter, more transparent version of Alexa or Google Assistant designed to serve as an “AI media navigator” finding me what I ask for in a savvy but unbiased way, and recommending me people and content that it genuinely thinks will be of interest to me. But this sort of vision is hilariously distant from the current reality in which the various Big Tech-controlled AIs are monitoring, modeling and guiding me with the objective of milking money and keeping me staring at certain particular apps or websites as long as possible.
One could argue that corporate control of our minds, hearts and bank accounts is the price we pay for all the valuable processing power Big Tech companies are utilizing to deliver us our daily diet of social media posts. But this is one more failure of moral and technological imagination. It is well within our current capability to deliver modern social media functions based on decentralized compute power as well as decentralized algorithms. This has been a theme in the blockchain world for years now, with projects like Golem, SONM and CPUCoin. SingularityNET has been incubating the NuNet.io project toward this end.
All of these decentralized-processing-power startups need some maturing to serve the needs of decentralized analogues of Facebook, Twitter and Tiktok. But the basic algorithms and architectures are there. What if decentralized, democratically controlled processing infrastructures had received even one-tenth of the financial and software-development firepower that centralized server farm infrastructure has?
With a decentralized infrastructure for social media, the whole issue of banning this or that person or topic from this or that social network would basically disappear.
Decentralized, transparently operated, secure, AI-driven reputation systems would rank and rate posts and posters, resulting in a system where nobody needs to happen upon posts they find offensive or disturbing.
On the other hand, if you don’t want to remain ensconced in your familiar community and idea-sphere and want to take the risk of contact with something new and potentially troublesome, you can tell this to your AI navigator as well.
The bad, the ugly and the glorious
Decentralization doesn’t cure all the perversities of human nature.
For sure, some people will use decentralized social media to plan crimes together. But people also gather in parks or bars for such purposes, and we don’t solve this by banning group face-to-face conversations out of earshot of government microphones. I have no doubt law enforcement will rise to the occasion of decentralized social networks as it’s done with prior tech innovations.
There is also no doubt that fraud attempts will be rampant on any decentralized social network and AI-driven navigators will need to rely on sophisticated “AI reputation system police” to militate against fraud and attempts to game the system. But the key point is this: The AI policing should be on the level of squashing reputation-system fraud, i.e., seeing through cases where people are pretending their content is something it’s not rather than squashing content itself.
Let us not forget that, alongside all the hate and stupidity, there is a massive influx of brilliant ideas and wonderful creations being put out there on the internet every minute of every day. Most of these have a hard time finding the audience they deserve because of today’s centralized online information architectures.
We have the core tech to enable radically more ethical and beneficial social media networks. All that remains is to get this tech out of the blockchain ghetto and into the internet mainstream. It’s not a small task but the potential benefit is huge. What we’re talking about here is not just a revolution in a certain sector of the software industry. It’s a massive upgrade in how the collective mind and heart of humanity guides its own growth. Personally I’m thrilled and a bit sobered to be in the midst of a tech ecosystem with so much potential to effect such deep and wide positive transformation.