Overview: Science&Technology

Wired's rules for using generative AI: "Not for images, yes for research, no for copyediting, maybe for idea generation". Meanwhile interactive book PromptCraft shows how to "get the most out of the creative man/machine communications... AI is like querying our collective intelligence".So how best to design an AI prompt?…

Noam Chomsky on ChatGPT, Bard and Sydney, which "take huge amounts of data, search for patterns in it and become increasingly proficient at generating statistically probable outputs — such as seemingly humanlike language and thought... hailed as the first glimmers on the horizon of artificial general intelligence ... surpassing human ... inte…

"Microsoft imagines helping clients launch new chatbots or refine their existing ones", using a version of ChatGPT trained on post-2021 content. "The service should also provide citations to specific resources... give customers ways to upload their own data and refine the voice of their chatbots... replace Microsoft and OpenAI brand…

"a modern version of del.icio.us meets Roam... a searchable, interconnected repository of the most insightful content on the Internet" - and a pretty interesting example of convergent evolution vis a vis myhub.ai

from Alice Albrecht, who "runs re:collect, a startup building an AI-powered thought partner:... advances in AI and cheaper compute lower the bar for getting from a creative idea to a final output... though, we still need to provide the initial seed... and ... judge whether we’re heading in the right direction. We’re still the creative direct…

ChatGPT-powered, the "‘My AI’ bot will ... initially only available for $3.99 a month Snapchat Plus subscribers.., eventually all... we’re going to talk to AI every day" - useful, if you're a messaging service.It's a "fast mobile-friendly version of ChatGPT inside Snapchat... trained to adhere to the company’s trust and sa…

While "Wolfram Alpha... seeks to distill scientific facts and perform calculations", ChatGPT and other similar AIs "build statistical models and string together sentences or pictures by calculating probabilities (of what the next word should be, for example). That has led to all kinds of mistakes ... Wolfram is not optimistic that t…

"What if we were to think of LLMs not as tools for answering questions, but as tools for asking us questions and inspiring our creativity? ... even simple tools can lead to interesting results when they clash with the contents of our minds"So he tries using ChatGPT as a muse. TL:DR; "ChatGPT asked me probing questions, suggested spe…

When James West "joined a small wave of users granted early access" to the new Bing (I'll call and tag it "BingGPT"), powered by the same LLM behind ChatGPT, which is "a great party trick ... powerful work tool, capable of jumpstarting creativity, automating mundane tasks", he soon "noticed strange inconsist…

Jeff Jarvis is not a fan of Ted Chiang's piece on CgatGPT, although he does admit TC (one of my favourite authors) "makes a clever comparison between lossy compression... and large-language models, which learn from and spit back but do not record the entire web". Indeed, he takes the metaphor further: "what is journalism itself…

Interesting, illuminating (but contested) metaphor for thinking about LLMs from one of my favourite authors, Ted Chiang:"Think of ChatGPT as a blurry jpeg of all the text on the Web. It retains much of the information... but, if you’re looking for an exact sequence of bits, you won’t find it; all you will ever get is an approximation... nonse…

The 3rd part of my 1/1/2023 bundle of 5 posts looks at how AI could turbocharge collective intelligence "and finance the resulting ecosystem, providing an alternative to Big Tech AI monopolies".

The "executive summary" of my 5-part bundle of 1/1/2023, which "provide a snapshot of my current thinking into how a decentralised collective intelligence ecosystem could be bootstrapped into existence."

Starts with articulating I've been meaning to write for many years about the oncoming AI-driven content flood: "The dark forest theory of the web: ... life-like but life-less state of being online... overrun with bots, advertisers, trolls, data scrapers, clickbait... algorithmically manipulated junk... eerily devoid of human life ... liv…

This piece captures and then explores exactly what I've always thought about the future direction of myhub.ai: imagine what would happen if you had your own personal AI assistant operating across your content - your public posts, private library (including content shared from friends, stuff in your reading queue, etc.) and the wider web, emph…

"Ipop gloop glog bluba droma floom gloope splog slopa" - This sentence means "The happy slime sees the earth under the sky while eating the slime's food" in Glorp. Here's the breakdown of the sentence:ChatGPT and I invent a fictional language spoken by slime-people... It understands subordinate clauses ... so understa…

Is ChatGPT creator OpenAI "a potential Google slayer. Why look up something on a search engine when ChatGPT can write a whole paragraph explaining the answer?" It's very capable, but "struggles to distinguish between truth and falsehood... often a persuasive liar... a bit like autocomplete on your phone... trained on pretty muc…

"WordPress developer Johnathon Williams ... demonstrated that, with a little bit of expert guidance, ChatGPT can drastically reduce the amount of time it takes to extend WordPress... the plugin [written by GPT] ... worked on the first try... the best results from asking ChatGPI to generate entire functions versus specific filters or actions..…

Another piece pointing out that ChatGPT, while "easily the most impressive text-generating demo to date... trained through a mix of crunching billions of text documents and human coaching", should not be trusted."a generative AI is what it eats", and ChatGPT ate countless biases in the content it processed. "No one should …

The claims re: ChatGPT by "fascinated evangelists... [that they] contain “humanity’s scientific knowledge,” are approaching artificial general intelligence ... even consciousness", are simply "a distraction from the actual harm perpetuated by these systems. People get hurt from the very practical ways such models fall short in deplo…

"ChatGPT is not overly reliable... surprising how often ... right, but they can also be plausible and wrong a lot of the time, and downright obviously wrong some of the time... responding with errors of fact, ... responses that may only be true for certain (biased) contexts... wary of providing evidence or citations ... [but] happy to make up…

Articulates exactly my problem with most Tools4Thought, hence the "super simple Thinking Tool" I want at the heart of myhub's CMS."Perhaps something as simple as having the word ‘Tool’ first causes us to focus on the tool more than the thinking... to confuse thought as an object rather than thought as a process... [causing us t…

"I spent a few minutes in conversation with ChatGPT."This was both me kicking ChatGPT's conversational tyres and exploring its limits, both those it admits to and those it does not.Key takeaways from a first reading:Many, but not all, content creators are screwedFor "writers, journalists, copywriters, consultants, and even publ…

"fine-tune GPT-3 on their own data, creating a custom version tailored to their application" improving reliability for more use cases and making the model cheaper and faster. Any dataset of any shape or size, plus "incrementally add data based on user feedback... less than 100 examples to start seeing the benefits ... doubling the n…

ChatGPT... is going to change our world much sooner than we expect, and much more drastically

Betaworx announcing "nine companies we selected and invested in for our THINKCamp accelerator program ... new technologies ranging from large language models and generative AI to graph databases and high-fidelity spatial user interfaces — to create tools for for how we may think."All probably worth checking out, ranging from Gordon Brand…

Discourse graph refers to an information model which represents scientific discourse by distilling "traditional publication down into more granular, formalized knowledge claims, linked to supporting evidence and context through a network or graph... knowledge claims ... as the central unit".Why? "Standardizing the representation of …

"I first decided to try scribing the Betaworx Render session on Plexus using Obsidian and try publishing it via GitHub (context). That didn't work, so when I saw [[rtnf md]] I thought I'd give it a whirl... the idea [of Plexus] is simple:post your thoughts into a defined shared brain on Plexus - eg : https://render.plexus.earth/the …

Another scribe of the Betaworx Render conference on Thinking Tools (context), this time in good old-fashioned blogpost form, with scans of handwritten notes to boot. Contrasting this with some of the other annotations, it's a good example of the difference between "someone else's notes" and "someone else's blog post&q;…

I've been invited to write a chapter for an upcoming book on Personal Knowledge Graphs (PKG). My chapter will encompass each user’s PKG, the Social Knowledge Graph created by networking them together via the Fediverse, Solid hosting, AI writing tools and Decentralised Autonomous Organisations.This post provides a first draft of its Introducti…