LovableSidekick comments on How it's gonna go

Funny

16,281 readers

6,236 active

1,915 users here now

General rules:

Be kind.
All posts must make an attempt to be funny.
Obey the general sh.itjust.works instance rules.
No politics or political figures. There are plenty of other politics communities to choose from.
Don't post anything grotesque or potentially illegal. Examples include pornography, gore, animal cruelty, inappropriate jokes involving kids, etc.

Exceptions may be made at the discretion of the mods.

founded 3 years ago

moderators

[–] LovableSidekick@lemmy.world 4 points 2 days ago (1 child)

"Hallucinations" are produced by LLMs, which wouldn't be appropriate for air traffic control. Automations of this type are done with Symbolic AI (aka "classical" or "logical" AI). Symbolic AI is built from sets of rules and conditions used by human beings to make the same decisions. This type of AI has been widely used in all kinds of control systems since the 1970s, and is completely unrelated to ChatGPT and other LLMs except by being called "artificial intelligence".

permalink

fedilink

source

parent

hideshow 1 child comment

[–] ErmahgherdDavid@lemmy.dbzer0.com 1 point 1 day ago

It'll probably be a deep learning model - some time series model that can factorise multiple input features like an RNN or LSTM. Those things work well within certain tolerances of their training data and are significantly more appropriate for this sort of thing than an LLM but as soon as something unusual happens they can react very weirdly - hopefully there are deterministic programmed guardrails that kick in and summon a human in edge cases.

permalink

fedilink

source

parent