ArtificialIntelligenceInToolsForThought

ThoughtStorms Wiki

Embedding failed with embed type :twitter

and data
{:type :twitter, :url "https://twitter.com/OGreenius/status/1592966200081846272", :title "", :caption ""}
java.lang.Exception: JSON error (unexpected character): <

Embedding failed with embed type :twitter

and data
{:type :twitter, :url "https://twitter.com/interstar/status/1592981291934580737", :title "", :caption "\n\nVery interesting.\n\nI agree with @codexeditor that there are obvious privacy concerns.\n\nBut also ... why do I want AI processing my writing in a TfT? Whose 'thought' is getting augmented by the tool? Mine? Or the AIs?\n\n\nFor a long time, I've believed that the greatest hits of personal productivity tools (including TfTs) were that they found a class of 'syntactic commonalities' but were 'semantically neutral' ie. an outliner processes tree-shaped data. A spread-sheet processes grid-shaped ... \n\nBut neither tells you WHAT you should be representing as a tree or grid. That's down to you and your individual goals. Nor does it care HOW you represent it beyond that basic 'geometry'.\n\nNow, though, we're gonna get AI tools that are VERY opinionated about your content \n\nAIs trained on huge data-sets from across the web are absolutely riddled with the biases and prejudices (And not just mean in the political / ethical sense) of 'everyone else'\n\nIf I use an AI-empowered TfT, are the thoughts it helps me have going to be constrained by that? \n\nThe point of a PKM / personal bicycle for the mind is surely to augment my own individuality and idiosyncrasies.\n\nIf, instead, it gives me some crunch of everyone else's individuality, what am I getting?\n\n\n\nNow don't get me wrong. I'm fascinated by the possibilities. I'm diving into AI Painting at the moment. And I'm wonderfully impressed by its creativity and potential.\n\nWhy don't I feel like this about writing?\n\n\nWell partly, of course, I'm totally untrained and incompetent with images. I love AI Painting because it gives me control over, and a way to experiment and learn more about, something I was previously ignorant and excluded from.\n\nIncorporating opinionated decisions \n\nof others is a worthwhile trade-off for me; for the chance to experiment and learn.\n\nPerhaps if I were already a visual artist, with a well developed aesthetic, I'd be far less excited by AI Painting\n\nMeanwhile, I flatter myself that I'm already pretty good at thinking & writing \n\nI'm not looking for a TfT to teach me how to think and write. Just to leverage my existing capabilities.\n\nIs this a valid intuition? Or is it just naivety? Would an AI that sorted incoming links and expanded scrappy notes according to its own criteria, nevertheless ... \n\nbe a genuine leveraging of my capacities?\n\n\n\n"}
java.lang.Exception: JSON error (unexpected character): <

Embedding failed with embed type :twitter

and data
{:type :twitter, :url "https://twitter.com/interstar/status/1592991933810749440", :title "", :caption "\n\nI guess my feeling is, if I write a short phrase and get a picture back, then it's obvious what I've gained. Because a picture is a worthwhile thing in itself.\n\nIf I write a couple of scrappy notes, x and y, and the AI goes on to infer a paragraph putting them together aren't all the ideas already there? Inherent in the notes?\n\nI guess it smells like it's automatically creating a kind of unnecessary extra verbosity / boilerplate. Whereas we OUGHT to be using TfTs to help us accept and work with raw nuggets of insight in the form of those scrappy notes\n\nGPT could turn these Twitter discussions into an essay\n\nBut I'd be happier with 'rollup' technology that just made it easier to capture, package and distribute them in the skeletal / concise form they're already written in. Which is already sufficient \n\nFor people to quickly scan and pick up the gist of.\n\n \n\n\n"}
java.lang.Exception: JSON error (unexpected character): <

Backlinks (1 items)

No Transclusions