[-] 8uurg@lemmy.world 7 points 1 month ago

Neat idea. This could be refined by adding a git hook that runs (rip)grep on the entire codebase and fails if anything is found upon commit may accomplish a similar result and stop the code from being committed entirely. Requires a bit more setup work on de developers end, though.

[-] 8uurg@lemmy.world 5 points 3 months ago

These models have so many parameters that, while insufficient to memorize all text it has ever seen, it can end up memorizing some of the content. It is the difference between being able to recall a random passage versus recalling the exact thing you need. Both allow you to spill content verbatim, but one is problematic while the other can be helpful.

There are techniques to allow it it 'read on demand', but they are not part of the core model (i.e. the autocmpletion model / LLM) and are tacked on top of it. For example, you can tie it search engine, which Microsoft's copilot does, and is something which I don't think is enabled for ChatGPT by default. Or allow it to query a external data bank (Retrieval Augmented Generation).

[-] 8uurg@lemmy.world 2 points 6 months ago

Do keep in mind that in Europe there are often required checks whether a car is 'roadworthy', in Norway this seems to be a biannual check: so you cannot really skip maintenance to the extent that that would be a huge factor.

[-] 8uurg@lemmy.world 13 points 7 months ago

I am not sure whether AI will make the creation of asset flips much easier though, given that store bought assets are already being dropped into a 'game' already.

[-] 8uurg@lemmy.world 1 points 11 months ago

You may be able to get more granularity by decoupling Google TV volume from the TV's own volume. To do this, you need to disable HDMI CEC control over volume in the remote's settings.

[-] 8uurg@lemmy.world 1 points 11 months ago* (last edited 11 months ago)

An hypothesis has to be falsifiable though.

[-] 8uurg@lemmy.world 4 points 1 year ago

KeepassXC is multiplatform. (Also: KeepassDX is quite nice as an an Android app for Keepass databases.)

[-] 8uurg@lemmy.world 5 points 1 year ago

For those taking this route: do note that the map data is downloaded from their servers. So don't forget to make a small donation to offset the server costs if you are able to!

[-] 8uurg@lemmy.world 9 points 1 year ago

It is already up and running, you can see posts from various government agencies at https://social.overheid.nl/public/local

[-] 8uurg@lemmy.world 12 points 1 year ago

It is uplifting though? The alternative is not 'not having cancer', it is 'cancer is recognized in a later stage'. Having cancer be recognized earlier really boosts the possibility of survival, and any improvement in that regard is quite uplifting in my opinion.

[-] 8uurg@lemmy.world 23 points 1 year ago

From the post of the account linked here (in Dutch): it is going to be a place for official government communication, not for individual government employees (and I presume, by extension, public registration in general)

[-] 8uurg@lemmy.world 2 points 1 year ago

There is a difference with adding an instance and accessing it that way (via their API - bypassing the server you signed up on) and via federation using the server you signed up on. Only in the latter case can you vote using the account already registered.

Try using the search feature to access the community via federation (e.g. "!cats@lemmy.world") rather than adding the instance directly.

view more: next ›

8uurg

joined 1 year ago