jeremylt boosted

Fun times we're entering, when Google can't bother to Google whether the bullshit emanating from their new LLM is true in the debut ad for that product.

From 2004:

jeremylt boosted

Join us for #EnzymeCon 2023 in Boulder, Feb 22-24! #Enzyme is the language-agnostic #LLVM automatic differentiation plugin. We're seeking contribution abstracts for:
* technical talks
* tutorials
* panels
* roundtables
* other formats

No prior experience with Enzyme necessary. If you could reimagine AD/differentiable programming, what would it enable?

jeremylt boosted

#libCEED 0.11 was released last week, providing performance-portable element algebra designed for integration in your code. These data structures move your sparse linear algebra from the dotted line to the solid line, giving a 2x performance boost even for lowest order elements (solid and fluid mechanics). This release has memory checking, backend and H(div) improvements, and various utilities, plus lots of new features in the implicit compressible viscous flow solver.

jeremylt boosted

It seems like Mastadon feels more complex than many people want, and providing a modifiable recommendation algorithm could easily deepen that perception. I think the expectations and needs of the open source software and academic communities (or any other specific community) are different than other users. User Experience design is a known hard problem :person_shrugging:

Oof, some people have some unflattering opinions of people who prefer a different presentation of data on their feed. 
I do not know the prevelance of this concern, but I've heard some people in my circle of friends say that recommendation algorithms can feel manipulative. TicTok seems to avoid this feeling (though the ads are staring to feel a bit manipulative). I wonder if providing insight into the recommendation algorithm and some simple model parameters to edit would help?

Character Prompt:
Freelance adventuring consultant

BBEG Prompt:
Evil megacorp continuing their ways after the apocalypse

jeremylt boosted

💡Are you a fan of #openscience and #foss? Consider volunteering to review for @JOSS_TheOJ! Open reviews are conducted via @github issues, and I have often had authors say it's the most positive/constructive review process they've experienced!


@ArielLeslie They really are disruptive without adding much good :(

Character Prompt:
Old school, pure support character

@likask @jedbrown
I've seen this discussion about the difficulty in learning to build and debug large projects in different contexts (freeCodeCamp, industry junior hires, ect). I agree that having our big packages in C/C++/Fortran makes this harder, but I assume that students find these skills hard to learn in Julia/Python/Rust based projects too. I feel like students who get to build a package are lucky - their skills scale with project size/maturity. That's hard to give every student though.

@jedbrown @likask
Its my impression that plenty of students get the gist of an abstract overview pretty well, especially with a good explainer, but then the leap from the abstract to the the implementation details is a big gulf to bridge. I've found 'war stories' helpful to a degree, but I feel like nothing is as effective as going on a bunch of bug hunts yourself.

Show more

freeCodeCamp's Mastodon instance.