All things are possible through Christ!

  • 0 Posts
  • 22 Comments
Joined 1 year ago
cake
Cake day: June 11th, 2023

help-circle

  • The term “AI bubble” refers to the idea that the excitement, investment, and hype surrounding artificial intelligence (AI) may be growing at an unsustainable rate, much like historical financial or technological bubbles (e.g., the dot-com bubble of the late 1990s). Here are some key aspects of this concept:

    1. Overvaluation and Speculation: Investors and companies are pouring significant amounts of money into AI technologies, sometimes without fully understanding the technology or its realistic potential. This could lead to overvaluation of AI companies and startups.

    2. Hype vs. Reality: There is often a mismatch between what people believe AI can achieve in the short term and what it is currently capable of. Some claims about AI may be exaggerated, leading to inflated expectations that cannot be met.

    3. Risk of Market Crash: Like previous bubbles in history, if AI does not deliver on its overhyped promises, there could be a significant drop in AI investments, stock prices, and general interest. This could result in a burst of the “AI bubble,” causing financial losses and slowing down real progress.

    4. Comparison to Previous Bubbles: The “AI bubble” is compared to the dot-com bubble or the housing bubble, where early optimism led to massive growth and investment, followed by a sudden collapse when the reality didn’t meet expectations.

    Not everyone believes an AI bubble is forming, but the term is often used as a cautionary reference, urging people to balance enthusiasm with realistic expectations about the technology’s development and adoption.















  • tl;dw

    • Cory Doctorow coins the term “enshittification” to describe how platforms start out benefiting users but eventually abuse users and business customers to extract all value.

    • Facebook started by prioritizing user privacy over ads but now prioritizes profits over all else.

    • Network effects are a double-edged sword - they lock users in but also make platforms vulnerable if users leave en masse.

    • Low switching costs due to universality and interoperability allow competitors to reverse engineer platforms and plug in competing services.

    • Mandatory interoperability and limiting data control can curb platform power by distributing control to users and smaller companies.

    • Recent antitrust actions aim to roll back decades of lax merger policy that let platforms consolidate power.

    • Breakups will take a long time so interoperability is a faster way to restore competition.

    • Laws should limit abusive behavior rather than rely on platforms to self-regulate.

    • Federated open services fail gracefully and encourage migration to better platforms.

    • Political will is growing but change will be gradual - focus should be on harm reduction in the near term.



  • This isn’t true, I think. You can have an instance that federates with nearly everyone but which still has a higher standard for behavior for its own users. This way, users on such an instance can see all the problematic instances but are not permitted to be problematic themselves. It’s an option.

    (Even still, I think you’d find yourself de-federating from someone eventually for spam or other technical reasons if not due to objections over content.)