• 1 Post
  • 14 Comments
Joined 1 year ago
cake
Cake day: June 11th, 2023

help-circle


  • Al Gore was definitely prescient in naming his documentary inconvenient.

    Climate change is as much a human problem as it is a geophysical one because that psychological defense mechanism that you anecdotally describe in the face of existential gloom is universal to our species, and the cause of so much ill-placed skepticism and hostility toward climate science and its communicators. Don’t Look Up also did a good job at portraying this unfortunate human bias.

    We as a species are too smart for our own good; smart enough to geoengineer our world to the point of threatening its existence, but not smart enough to address our own resistance to change and take collective action where and when it’s urgently needed.

    For those who study climate change and those who try to mitigate it, there is this double burden of not only losing sleep over the magnitude of the existential threat, but also facing the moral and psychological failings of those who refuse to see reality for what it is and argue against it. It’s tiring.







  • I agree with you.

    A company’s core business and skillset is rarely to manage an on-prem IT infrastructure, which is a highly complex endeavor these days. Security most always benefits from being put in the hands of cloud providers such as Microsoft, Amazon, or Google, who can mobilize the best talent and apply economies of scale and modern best practices to cybersecurity across an entire stack.

    It also means far fewer liability headaches for the companies that transfer this difficult and onerous responsibility to cloud providers. It’s not even necessarily cheaper to go full cloud; I’ve seen multiple examples where it wasn’t, but the reduction in complexity and liability made common sense. So even the “LaTe-StAgE CaPiTaLiSm!!” claim is just a tired trope at this point.

    It’s easy to focus on one publicized exploit of Microsoft’s cloud like this one, and not see the other side of the argument of how many exploits were avoided over the years by not having individual companies manage their own servers. It’s still entirely plausible that the general move to cloud infrastructure since the late 2000s is a net win for cybersecurity in aggregate.

    I would also add that whether other cloud customers might be breached simultaneously in the extremely rare event of a cloud-wide exploit is not a consideration when a company decides to move from on-prem to cloud. It’s just a Moloch problem that doesn’t and shouldn’t concern them.






  • I agree that investors requiring demonstrable returns has played a role in this cycle. Steve Huffman is desperate to show profits ahead of Reddit’s IPO, and Musk is desperate to recoup his $44B investment in the blue bird.

    However, I believe that there’s also another consideration. Many of today’s platforms started out with a somewhat idealistic intent. Jack Dorsey wanted Twitter to be an open protocol, though never quite achieved his vision. Aaron Swartz contributed to the open design of early-days Reddit. Facebook was meant as a non-profit university community builder. Google had (and abandoned) a “do no evil” motto. Etc.

    The original user-first approach of these platforms created organic growth and encouraged ambassadorship by motivated users who became frequent contributors, unpaid moderators, etc.

    Over time, however, people moved on (Dorsey, or very sadly Swartz) or got greedy from success (Huffman, Zuckerberg). The focus shifted from user-first to advertiser-first. Platforms like Reddit still used a loss-leader approach of losing investor money on frills such as API because it helped sustain growth for a while longer.

    But once critical mass was reached, there was no longer a need to coddle the most enthusiastic and long-time users. They had exhausted their usefulness. The platforms could finally embrace the advertiser-first model in which the user, not the content, becomes the product.

    So here we are with the worst of both worlds. Reddit could have offered a reasonable paid API plan that would have allowed the thriving third-party ecosystem to retain the power users and contributors. Instead, it went all-in with a walled-garden approach buoyed only by advertising money, even if it means that the content quality dwindles. Twitter also went “private” in the sense that an account is now required to even view the content, and aggressively promotes its paid plan to users –who are still subject to interstitial ads and promoted content– even for basic hygiene features such as 2FA.

    As for why Reddit, Twitter, and Discord shit the bed at almost the same time, part of it has to do with VC pressure (as mentioned by the parent), and part of it is they are the same generation (more or less) of social networks and are reaching an equivalent stage where buyout (Twitter) or IPO (Reddit) is the next logical step.

    The writing is on the wall that a paradigm shift is in order. The pendulum has considerable momentum, though, and will allow the centralized, walled-garden web to thrive for a while longer, just like Facebook survives catering to mostly an audience of unsavvy boomers. But the swing back will gradually enable alternative models to grow that are based on open platforms and federated content. We’re just very, very early in this cycle.

    Oh, and sorry for the long-ass essay, I got a bit carried away.


  • I’ve been online since circa 1993 and for the first decade or so, discoverability was a challenge due to the lack of efficient search engines like Altavista or (later) Google.

    Webrings consisted in individual website owners (e.g., on Geocities) placing one or more banners at the bottom of their webpage linking to other like-minded sites, typically in quid-pro-quo manner (I link to you, you link back to me), or to a manually-curated directory of like-minded sites.

    This was when “surfing the web” meant exactly that - you would surf from one site to another using hyperlinking within web communities. Bookmarking was then how you kept track of the most interesting sites you came across.

    Now there is hardly a need for hyperlinking and bookmarking, since much of the content is centralized on a few platforms, and search engines take care of the discoverability of niche content.