Stephen Wolfram presents an in-depth look at the just announced Wolfram Notebook Assistant + LLM Kit.
Originally livestreamed at: [ Ссылка ]
Learn more here: [ Ссылка ]
Stephen's announcement blog: [ Ссылка ]
00:00 Start stream
2:35 Announcement + Demo
1:08:40 Q&A Begins
1:08:53 Innovative tool for data visualization. Enhances workflow efficiency. Bridges computational tasks seamlessly. How do you envision using this in your projects?
1:10:19 How does that contrast with other notebook assistants, particularly in terms of operating in terms of a computational language (WL) vs natural language?
1:12:22 Which model does Notebook Assistant use 'under the hood' by default? Can you choose others?
1:13:30 How is this different than Chat Notebooks?
1:15:33 Most of us don't dream and think in WL terms as SW does, so that's quite helpful to bridge, and the more powerful and integrated the assistant the better.
1:16:31 Can you ask it to make a bifurcation diagram to examine the onset of chaos in something like an RLC circuit?
1:18:57 Is it possible to give some insight into the underlying computing load and infrastructure, and scalability and such?
1:19:45 Could it generate random origami 3d objects?
1:20:54 So far what are your favorite integrations between the assistant side-chat and the WL notebook itself, versus another generic LLM in a side window and such?
1:21:33 Are there plans to make it working in other languages?
1:23:47 Has the Assistant had any, "happy accidents," so to speak that were particularly useful?
1:25:10 Make a graph showing a breakdown of a trig identify, including alternate transforms, expression in each bubble in traditional form.
1:31:23 Create a slice of pizza using Gaussian curvature.
1:34:15 Could I add all my (hundreds) of notebooks to its semantic base to incorporate into responses?
1:35:17 What about a natural language exploration of a semi-complex dependency graph as one would find in a modern software artifact (or even just using WL functions from the repository as a stub here)?
1:36:48 Can you try this prompt: Create a tool that generates cellular automaton-based snowflake patterns, including a GUI with real-time sliders to adjust parameters like rule number, grid size, and speed.
1:39:36 Can it accurately translate phrases into propositional logic expressions and output their truth tables?
1:41:51 Wrap up
1:43:52 End stream
Follow us on our official social media channels.
X: [ Ссылка ]
Facebook: [ Ссылка ]
Instagram: [ Ссылка ]
LinkedIn: [ Ссылка ]
Stephen Wolfram's Twitter: [ Ссылка ]
Contribute to the official Wolfram Community: [ Ссылка ]
Stay up-to-date on the latest interest at Wolfram Research through our blog: [ Ссылка ]
Follow Stephen Wolfram's life, interests, and what makes him tick on his blog: [ Ссылка ]
