Peter Goldstein

Very soon, the blocker to using AI to accelerate science is not going to be the ability of AI, but rather the systems of science itself, as creaky as they are. The scientific process is already breaking under a flood of human-created knowledge. How do we incorporate AI usefully?

Amusing how 99% of people interacting with reality forget how this thing works. It's an advanced extremization machine. It generates the next instant of time based on the Cauchy surface and the action. Under the hood, it's a giant volume integral that has eerily good output.

Working on Torch 🔥 Unified health record + LLM in one iOS app. Syncs records from hospitals, labs, Function, One Med, PDFs etc. Makes it simple & fast to get all your health data in one place + get LLM help making sense of it. Reply/RT for TestFlight invite. More below ⬇️

@emollick If I can make an offbeat suggestion - Douglas Adams. His AIs have personalities that are more important than their capabilities. Think of Marvin. His vision for society and meaning is a little harder to suss out.

Very long traditions of AIs trying to find mesning themselves (Leckie, Murderbot & many more), and in dealing with cold logical AIs that know everything (Asimov and many others) or post-scarcity worlds without serious AI (Star Trek), but not a vision of an LLM future.

Apropos of Pantheon, it is interesting how few science fiction novels explored finding human meaning in a non-dystopian world with advanced LLM-like conversational AIs (But no AI main character) Banks' Culture is obvious but who else? Becky Chambers? Early Stross? Egan? Doctrow?

We know how to assign responsibility for automated systems, many decision-making systems are already automated in modern firms (it is why the classic IBM sign never made a lot of sense to me.) What is much harder is the wider question of who is change of organizational change.

@railsui_ Is it listed on Context7? How easy is it to use with Gen AI dev tools?

📍Have requests for themes or components? I'm all ears.

Tired of pasting similar-looking Tailwind UI bits into your Rails app and hoping for the best? I was, too. Enter Rails UI — a real UI system for Rails devs. 🧵👇

As an aside: if tech could export one cultural quirk to the rest of the world, it should be the culture of end to end ownership and agency, ideally throughout the chain and in the avatar of the founder if we cannot manage that.

@joemasilotti Got it. Thanks.

@petergoldstein It depends on your timeline. If you're fine pointing to v1.1.x then you can use the book as-is today. But if you have the time then waiting for the next release might be better. The biggest change will be the chapter on tabs. Everything else falls line with the new approach.

I'm reworking my ENTIRE BOOK to work with Hotwire Native v1.2, released yesterday. 2.5 hours in and I've completed… one chapter. Oh boy! This might take a while. 😅

@aollivier82 @scottlincicome Our family moved from SF two years ago. Moving to Raleigh has been an unbelievable upgrade.

@scottlincicome As someone who moved from LA to Raleigh, it feels like I moved from the third world to the first. Turns out, clean streets and low rent are a choice!

Related:

We hope MIT will follow Harvard's lead on institutional neutrality, because it's good in itself, but also because, frankly, we'd like to be out of the position of telling MIT it should be more like Harvard as soon as possible. Just not a natural state for us.

@peterrhague Potentially a little dated, but you might consider the Heinlein juveniles - en.wikipedia.org/wiki/Heinlein_…

Is there a science fiction equivalent of Harry Potter? My ten year old is very keen on sci fi but a lot of the classics I’d recommend are a bit above his reading level.

@llama_index @metaphorsystems Will web search work with OpenAI on Azure?

Introducing RAGs v3 🌟: Build a RAG bot that can also search the web 🌐, to find answers that aren’t immediately in its corpus. Do this all in natural language, not code! 💬 Get a better experience than ChatGPT + Bing ⚡️ We used our integration with @metaphorsystems - a search engine designed for LLMs - to pull in relevant text from the internet. Additionally: - you can now view the tools that the agent is using. - note: web search is only accessible for our OpenAI agent Check out the video below, and get started with the repo today. Repo: github.com/run-llama/rags

@emollick I was surprised to see this competition pop on on Kaggle today as a Featured Code Competition - kaggle.com/competitions/l… . People are having a very hard time with the idea that LLM generated text is undetectable.

AI work is ultimately undetectable, despite the recent discussion of watermarking. AI writing is undetectable by any automated system after just a few rounds of prompting or revision This paper shows it is also easy to defeat watermarking for AI image. arxiv.org/pdf/2305.03807…

@jnunemaker Ruby's "convention over configuration" standard makes Copilot really, really good at Ruby. Most Ruby projects have very similar directory and code structure, so it's much easier to generate "standard" code. Both the code completion and the chat give me great answers in Ruby.

First time using GitHub Copilot chat. I asked how to use def_delegator in Ruby. Got perfect example usage w/o leaving editor. I wanted to rename the delegated method, so I asked "how do i rename the delegated method when using def_delegator" and it just worked. 🤤

In 2006, I was 1 of 4 designers on Google Search. For 20 years, every search engine has copied Google. Now ChatGPT, Bard + Claude look like Google's offspring - "better” search engines. But last week signaled we're on the brink of a design revolution. ChatGPT unveiled incredible new features. These could give us the opportunity to completely shift how we interface with AI. Here's the full story: ––– When I was a designer on Google Search, all major search engines looked the same – Google, Yahoo, MSN Bing. Google was the market leader with a heavily optimized UI that supported billions of dollars in ad revenue. Naturally, it became THE way to show search results. Its success made it illogical for Google to consider big UI changes. And any changes they did make were just mirrored by everyone else. So 20 years later, we’ve only seen incremental changes to search engine UIs. ––– Today, we have consumer-ready LLMs (Large Language Models) freshly in our hands. As consumer products, these are in their infancy. We’re very early in understanding their capabilities and defining how people interact with them. These are uncharted waters. And yet ChatGPT, Bard, Claude etc. all chose a text-based input box — just like Google’s search box — as the core interface. Why? The input box is simple, versatile, and familiar. - It’s simple to understand → you type your questions into the box. - It’s versatile → the box can handle all sorts of questions/queries. - The paradigm is super familiar → people immediately know how to use it. Because of this, LLMs have essentially become “a better Google.” ––– But last week’s ChatGPT announcements thrust open the doors to new possibilities. ChatGPT is now multi-modal — it can see, hear, and speak. These are the recent announcements from @OpenAI : Voice: x.com/openai/status/… Photos: x.com/openai/status/… The example of ChatGPT explaining how to lower a bike seat was incredible. But, it could be so much better! The video showed you'll have to post multiple new photos to keep adding new information and to progress the conversation. It was still a linear conversation centered around the text box. But what if we rethought the interface to center around the image? What if ChatGPT supported both images AND voice simultaneously? Could we end up with a more immersive experience? ––– How else could interacting with LLMs mimic IRL conversations? Could we (or the AI) pinch to zoom or rotate the image? Could we interact in real time with video? What new possibilities open up with context being preserved over time? ––– There is so much energy and excitement around what AI can do. But we are limiting the potential by assuming the conversation box is the best interface. Right now, designers have the chance to create truly novel interactions and bust through the 20+ year old search UI paradigm. The ideas above are just to illustrate some potential options. But they are also intended to spark a flame. Now is the opportunity to be creative and explore divergent UIs. What are the craziest, coolest, most creative UI ideas we can unleash? LFG 🚀

And here's the story - baen.com/chapters/W2000… (5/?)

How we answer that question likely changes our views on what content should be used to train generative AI systems. (4/?)

Had a number of great conversations about generative AI at yesterday's Fortune #CEOInitiative event. During one such conversation I was reminded of an old science fiction story (Melancholy Elephants by Spider Robinson) that I think has relevance for today's debates 🧵 (1/?)