Modifying and adding noteworthy

Author: AramZS

public/img/posts/chatgpt-answers-motivated.jpg

@@ -10,6 +10,7 @@ tags:
 series: The Wild Web
 group: The Wild Web
 cover_image: 1443px-Typical-orb-web-photo.jpg
+featured: true
 ---
 
 > When I envision the web, I picture an infinite expanse of empty space that stretches as far as the eye can see. It's full of fertile soil, but no seeds have taken root. That is, except for about an acre of it.

src/content/essays/2024-06-08-the-internet-is-a-series-of-webs.md

@@ -11,6 +11,10 @@ cite:
 
 ---
 
-Here in the scam economy, we have more proof that capital is ever more disconnected from anything of true value or any of the labor involved in generating it. The rise of crypto currency in the immediate post-election period makes it clear that you absolutely can bribe the market, buy the government, and enrich yourself without consequences. There was never a true laissez-faire economy, but there definitely isn't one now. The economic conditions we've created post-2008 are: the biggest successes are only available when you make yourself too big to fail and then fail hard. The systems of government and investment will continue to poor money into your venture and, in your failure, you have to do less and less of any true value. This means you can extract ever more value upward to the very top, confident that--once you hit a certain size--you will never feel any consequences of your actions; those consequences will be felt by everyone else. We live in the moral hazard of the billionaire class. In the scam economy you want the line to go up until finally it creates its own perverse gravity, drawn up forever regardless of what might happen to briefly push it down. 
+Here in the scam economy, we have more proof that capital is ever more disconnected from anything of true value or any of the labor involved in generating it. The rise of crypto currency in the immediate post-election period makes it clear that you absolutely can bribe the market, buy the government, and enrich yourself without consequences. There was never a true laissez-faire economy, but there definitely isn't one now. 
+
+The economic conditions we've created post-2008 are: the biggest successes are only available when you make yourself too big to fail and then fail hard. The systems of government and investment will continue to poor money into your venture and, in your failure, you have to do less and less of any true value. 
+
+This means you can extract ever more value upward to the very top, confident that--once you hit a certain size--you will never feel any consequences of your actions; those consequences will be felt by everyone else. We live in the [moral hazard](https://aramzs.xyz/glossary/moral-hazard/) of the billionaire class. In the scam economy you want the line to go up until finally it creates its own perverse gravity, drawn up forever regardless of what might happen to briefly push it down. 
 
 > The fundamental value of one bitcoin — the amount of economic value a bitcoin can directly generate — remains zero dollars. But the speculative value of one bitcoin is one hundred thousand dollars, because that’s what the marginal investor has agreed to pay for it.

src/content/noteworthy/2024-12-06-bitcoin-is-worth-dollar100000-it-still-doesnt-have-a-single-legitimate-use-case.md

@@ -0,0 +1,23 @@
+---
+title: Stop using generative AI as a search engine
+tags:
+  - list/in-the-news
+cite:
+  name: Stop using generative AI as a search engine
+  author: Elizabeth Lopatto
+  href: >-
+    https://www.theverge.com/2024/12/5/24313222/chatgpt-pardon-biden-bush-esquire
+
+---
+
+This article is great at noting how ridiculous and harmful the framing of products like ChatGPT. I also noted on [Bluesky](https://bsky.app/profile/chronotope.aramzs.xyz/post/3lcih4gj4w32w): 
+
+> Watch this "magic". If you frame the question to get what you want, it gives it to you, true or not.
+
+![ChatGPT giving motivated answers](/img/posts/chatgpt-answers-motivated.jpg)
+
+The thing about ChatGPT (and the rest of the generative AI systems) is that it is basically--at a fundamental level--complicated autocomplete. It isn't giving motivated or biased answers, it is just looking for the most likely completion text for your question which, unlike news headlines, often gives you the answer you want instead of the answer that is correct. This isn't just a design choice for ChatGPT (though it is partially that) but also a result of sites building pages around SEO goals for more than a decade. It used to be that the bad--but well SEO optimized--answers would be excluded by the weight of links to the right answer and moderation action against fraudsters and checks against exploitative behavior by site builders. ChatGPT has none of that data involved in how it responds, it just finds the words in a multidimensional space closest to yours by drawing a line through its massive data set. That massive collection of data is a scrape of the whole web and therefore includes all those fraud sites built to try and get your click. 
+
+By the way, this problem is also one of the many reasons why Google Search increasingly sucks now, the algorithm for sorting results has increasingly leaned on click-through and time-to-return-to-search-page as value signals for site. That means that a site that gives you a concise, simple, correct answer gets devalued compared to a site that forces you to click through 5 ads to get a self-satisfying distraction. Like GenAI, modern Google Search uses human behavior but fails to account for or understand it.  
+
+> Technology that actually serves people takes into account human behavior. Maybe there is a way to make generative AI useful, but in its current state, I feel tremendously sorry for anyone gullible enough to use it as a research tool.