GPT dot showed up in my European Pro account overnight, but it's asking me to upgrade to a new Pro plan to use it at the same cost. Is this a plan with lower usage? Has anyone tried?
Jens Bosse
@bosse_lab
Illuminating virus dynamics @CSSBHamburg through @MHH_life and @RESIST_cluster. A little zoo at home. Equal responsibility dad of three. Views own
486 Following 905 Followers 172 Posts
Latest posts 5
Try ASD-STE100 to make Opus 5 in Claude science to cut the cr*p! The best way I found so far to make Opus 5 understandable
We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:View quoted post on X
Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:
Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:
Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:
Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!
In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
Fables safeguards are hilarious. Asking about the contents and historical context of a 100+ years old postcard must be a cybersecurity or biosafety issue… 🤣
Me to Fable on an empty repo: "Let's come up with a fun project that does not get stopped by safety guards." Fable: "Fable 5's safety measures flagged this message." 🤣
On June 8 at 19:00, CSSB will be hosting a panel discussion entitled “Coding Life: Leben gestalten mit biologischen Werkzeugen.” Registration: cssb-hamburg.de/codinglife