Tootfinder

Opt-in global Mastodon full text search. Join the index!

No exact results. Similar results found.
@mela@zusammenkunft.net
2024-10-23 17:17:36

Was mich an der Sache noch besonders abfuckt … die minderjährige Tochter des Fahrzeugführers war mit im Auto. Nicht nur ein Leben zerstört, sondern auch noch dem eigenen Kind ein Trauma verpasst. Wie kann man sich so sehr nicht im Griff haben, als erwachsene Person?
mastodon.social/@tazgetroete/1…

@austegard@mastodon.social
2024-09-25 21:04:32

Molmo (molmo.allenai.org/) might have its talents and a place, but reasoning is not one of them

User, showing a picture of four die, with two of them fairly vertical (Blue showing a 2, Red showing a 5) one diagonal and one obscured:

What are the values on the sides of the die facing up?

Molmo:
The values on the sides of the die facing up are:

Blue die: 6
Red die: 4
Green die: 5

These numbers represent the top faces of each die in the image.

==> Entirely fabricated, inconsistent and incorrect
@toxi@mastodon.thi.ng
2024-10-14 16:26:08

Old graphical editor prototype (2019/20) with various ways of managing multiple selections of graphical entities, incl. align/distribute, duplicate, invert, select/deselect region (box/convex hull). editor is generic. items could be files, shapes, mesh vertices, 2d/3d points, graph nodes, whatever...
Been re-animating this prototype and toying with integrating it into my personal note taking & media management tool to organize items into piles for batch-processing and to easier app…

Screenrecording of a graphical editor prototype showing a number of circles being duplicated/manipulated and re-arranged via a sequence of mouse gestures and keyboard commands (the most recently executed command is shown in the top-left corner). The steps are:

- turn on grid
- create/duplicate circles (with grid snapping)
- selecting all
- duplicating selection(s) to form an 8x8 layout
- manually selecting 4 circles and then all circles in the convex hull
- drawing multiple box selectio…
@austegard@mastodon.social
2024-08-19 16:08:18

Not sure how universal this is but I simply have no time or energy for "dumb" LLMs like Claude Haiku, or most open source models -- they are firmly in the Annoying Valley of pseudo-intelligence.

The diagram illustrates the concept of the "Annoying Valley" in AI development. It shows a graph with two axes:

1. The horizontal axis represents the breadth and depth of AI capabilities, ranging from narrow AI to AGI (Artificial General Intelligence).
2. The vertical axis represents user satisfaction.

The graph depicts a curve that:

1. Starts low on the left (narrow AI capabilities)
2. Rises steadily as capabilities increase
3. Reaches a peak
4. Then dips into a trough labeled "The Annoying…
@kronomoonphotos@universeodon.com
2024-08-18 23:05:39

#LibreTranslate
"red #moon tonight"
Enormous, gorgeous, brilliant #rose #gold near-full #MoonPhoto from our #foto #friend Arthurr :debian: ⏚ (@…).
Hey PieR @… thx for the #poke! 🤗
🔗 #SturgeonMoon #KronoMoonPhoto #MoonOverMastodon #FullMoon #Lune #Luna