Web2 is just full of bots. Will Web5 fix this?
Humans no more are the majority on the web, as the latest stats shows. And that’s even though there are 5 billion internet users globally. Yet we account for only 38.5% of internet traffic, as per World of Statistics. Seems those digital white walkers are pushing us out of the web (and having Elon Musk in a tizzy).
It’s common knowledge that Web2 is overpopulated with bots, e.g., social media are full of bot accounts and comments. But isn’t 61.5% of non-human traffic an exaggeration? OneArt delves deeper into the issue to find out if these numbers have their share of truth, whether we’d better get ready for the bot takeover, and what a solution can be.
Is bot winter coming?
According to the Barracuda research results, it’s too late to get prepared since bots already make up 64% of traffic. North America is leading in terms of regional distribution (67%), followed by Europe and Asia.
And the icing on the cake. Bad bots follow a standard workday (well, at least they do).
However, there are more optimistic studies. Imperva Bad Bot Report shows that bad and good bot traffic share is 42.3%. But there’s a negative trend: the volume of bad bot traffic is soaring, e.g., during the holiday shopping season in December 2021, it made up 30% of all traffic. Moreover, most of it is generated by evasive bots that can bypass common defenses.
Well, bots may outperform humans, so what?
You may not know that you deal with bots when commenting on your favorite NFT artist’s posts or sharing them. Moreover, creators may have no idea of what’s going on with their accounts. That’s how engagement farming works.
Some bots are “more equal than others,” enabling misuse, stealing the data, causing breaches, skewing website performance, and carrying out other malicious activities. For example, an account takeover — a type of identity theft. Last year, ATO attacks surged 148%, with financial services the most targeted.
Stolen credentials are just the beginning, often followed by fraudulent purchases, money transfers, phishing campaigns, or warranty fraud. And more account takeovers since 66% of users have the same password for various accounts.
Are social bots top influencers?
The scale of the disaster may be more significant than your favorite FB profile stolen. Social bots can spread misinformation and conspiracy theories, tarnish reputations, manipulate stocks, and inflate support for a specific candidate. Twitter bots are said to have affected the 2016 U.S. presidential campaign and were active during elections in France and Germany.
“About one in every five election-related tweets from Sept. 16 to Oct. 21, 2016, was generated by computer software programs called “social bots.”
Emilio Ferrara, Research Assistant Professor of Computer Science (University of Southern California)
Bots managed to gain a large base of followers and influence. At that time, Twitter estimated their share at 5–8.5% of all accounts.
In 2022, Twitter bots — or rather their number on the platform — wrecked a $44 billion deal to sell the service. In short, Elon Musk doubts Twitter provides accurate numbers of spam bots in its user base. The company has sued the billionaire to force him to complete the deal.
Infamous Russian bot farms have been working at full throttle for years to spread propaganda. Alleged of interference in elections in western countries, they now distribute fakes to sow panic, spread misinformation online, and justify the war in Ukraine. As well as facilitate attacks to disrupt critical infrastructure.
How Web2 tries to fight bots
Bots adapt to detection and defense like bacteria adapt to antibiotics. They exploit human biases and become harder and harder to identify. For instance, bot-generated tweets are shared at the same rate as those posted by humans.
“You can never be 100% sure whether a profile is a bot.”
Simon Hegelich, political data scientist (Technical University of Munich)
To distinguish bots from real people, experts suggest:
- Using machine learning
- Implementing biometric authentication
- Using graph-based, crowdsourcing, or feature-based detection. Or combining them (learn more)
- Monitoring your traffic sources, spikes, and failed attempts to log in.
However, there are no signs the issue can be solved once and for all. Rather the opposite — bot traffic volume will increase and remain a pain point. At least, we probably miss most of the advanced, AI-based bots so far, Emilio Ferrara says.
Aren’t you fed up?
How we can respond
DIDs are self-owned and can be generated by individuals or organizations via a system they trust. Authentication can be provided with cryptographic proofs, e.g., digital signatures.
VCs can provide all the data that IRL credentials provide but in a more trustworthy and tamper-evident way due to digital signatures and blockchain-based nature. (Doesn’t that sound familiar to you? Check out our Soulbound NFTs Explained article.)
🔥Context management in a so-called identity wallet can grant users true control over their data, allowing them to specify, maintain, and enforce how other people, services, and apps access and use their identifiers. Guaranteed identity portability would be a go-to choice.
Is Web5 realistic?
The Web5 concept was met with both enthusiasm and criticism. Some point out that a Bitcoin-based system limits the opportunities, excluding projects built on other blockchains. Others doubt it can be implemented in the foreseeable future and scaled to kickstart mass adoption.
As for the latter, we at OneArt think that adding social features to the identity wallet mentioned in TBD’s presentation could help solve the issue. Making the wallet an all-in-one solution for various user needs, improving UX, and helping people not only manage their assets but also a like-minded community is the key.
Of course, we won’t find ourselves in a “bright future” of Web5 either tomorrow or in a month. But we have at least a sort of draft of a roadmap to reach it and throw off the yoke of overbearing monopolists.