Hacker News Reader

Comment by KittenInABox | Hacker News Reader

More and more I feel like online spaces are far more insidiously dangerous than anticipated. We have bots or malicious actors among the userbase explicitly to sow dissent and spin the quality of discussion downwards. We have algorithms that push content based on engagement, which isn't actually connection or anything resembling engaging with community. We then have online cultures that encourage lowering empathy for others and reveling when bad things happen to people who "deserve it", like "leopards eating faces" communities or punching women or whatever. I'm not being a centrist or anything like that but I feel online spaces actually socialize people to be lonelier by encouraging engagement via positions and activities that actively isolate us from each other, as opposed to genuine connections that come with say community service IRL.

View on HN

I think the next decade will include a shift where we acknowledge that the web is a dark forest and we stop trying to block the scams/propagandists/malware/ads specifically and instead we start blocking everything that isn't signed by somebody we've met in person and trust explicitly (or somebody that they have met in person, etc...).

We're going to have to learn a new kind of hygiene.

That's already here for me, and others I believe as the most interesting communication has moved to group chats. The open web is for advertising, and the best LLMs will be for information retrieval.

I'm a bit doubtful. You're seeing my words... do you have data on the transitive trust path between you and I? In the world I'm describing your browser wouldn't render this text unless it had confirmed that path decided that it was acceptable.

I'm exaggerating a bit, but the broad outline is essentially there already I think. The move to walled gardens everywhere (if we call this site a walled garden) is the first move almost everyone has made in terms of where they allocate attention. I think people will just move to smaller and smaller circles of interactions they trust (discord/group chats) especially if you don't need to query the open web for information anymore, and you trust an LLM agent for information retrieval / commerce.

I don't trust LLMs with the truth. I think they are an amazing tool for !!!generating!!! text; but they're wholly untethered from any sort of ground truth. Unfortunately what will happen, and I hate being right about, is that LLMs on everyone's phones will give dumbasses their own personal conspiracy/radicalization machine. You can generate all the text and images you'd ever want to give yourself the thrill of hate without any tether to reality.

Except that phishing means our trusted contacts will get popped...

Trusting people with poor information hygiene is a hazard. Tell your friends. Train your parents.

I don't get how or why it seems to have gotten worse in the last 20-25 years.

I was a very online teenager at the end of the 90s and I have formed friendships that last to this day. Or at least staying in touch. And it's not just me, it's the same for a lot of people I got to know offline who have different such circles.

Maybe because it was just a smaller circle of people? Maybe because it was so not normal that people were much more cautious? Or because half of the online venues were run by hobbyists? Or because no companies and bots and people trying to make money were involved. Probably the latter.

See https://en.wikipedia.org/wiki/Diffusion_of_innovations

Early adopters of just about anything are smarter and better than the people who come along later.

IMO back in the 90s anyone looking to do harm on the internet was working alone or in a small group. That's bad, but not a system-level danger. It wasn't trillions of dollars worth of market cap conspiring against us.

I'm also a little confused, because I remember such common empathy-lacking rhetoric like demanding boob pics of any proclaimed female on the internet being common in my tween online years. Or sending shock stuff as a joke (2girls1cup, tubgirl etc). So its not like the internet didn't have troll-y cesspits. And yet somehow something has changed such that as a population we've been trained against forming human connections. Even 4chan organized against scientology once upon a time, that required at minimum real organizational labor and forming some kind of community action. I can't imagine any similar thing happening now.

paypal brought us freedom.

I think it's more dangerous as a parasocial playground. You think you are making friends, but in fact it's a poor substitute and doesn't carry the risks/benefits of proper bonding with another human. This is independent of whether the other person is a bot or their communication is curated by an algorithm.

Can you say more about those risks and benefits?

Risks - you might be exposed to an idea that challenges you, you might be bored for stretches of time, you might think the relationship is stronger than it is for the other side and embarrass yourself, especially if you read romantic interest. You might have to alter your behavior in the name of social grace. You might lose that companion.

Benefits - IRL human companionship.

A system is kind of like an unthinking lifeform. It will adapt to find ways to make itself stronger and to grow. It will use whatever resources it has access to to and does not care about (or is even conscious of) the damage it does in the process. All systems are like a virus in this way. Social media as a system is just playing this out. The social media system feeds on engagement and use and, it turns out, that isolation and polarization lead to more social media engagement. So it adapts itself to make more that, the users well-being is irrelevant. Corporate bureaucracy, capitalism, religion even are similar. They may have been created for some purpose (that was useful for the people who created it), but once they reach a certain scale, that purpose inevitably shifts to simple self-perpetuation.

Another example is a charity formed to help with some problem which it initially does well at. But then over time the charity matures and more and more of its funds are devoted to keeping the charity running, paying salaries, renting office space, paying for trips, organizing fund-raising events. Less goes to the cause. The primary purpose of all the fundraising then becomes to cover the overhead of doing all the fundraising. The cause is a second order consideration at best.

Social media has matured from being about connection, information exchange, community building. Now its just to keeping the clicks and comments coming so the ad-revenue and data-mining can be used to cover the cost of the servers, developers, and executive bonuses.

Reply on news.ycombinator.com