PoisonedPrisonPanda @ PoisonedPrisonPanda @discuss.tchncs.de

Posts

3

Comments

455

Joined

2 yr. ago

die hitz geht immer.

how does LLM training benefit from metadata?

thx. i like that

edit: laughs in german

• for syncthing. it aint a backup solution. but it makes life simpler

oh yes. please feed my anxiety.

simply encrypt the files before sending?

I mean that calling deepseek out for shady privacy practice is fine, but by open sourcing their model they are at least better than most american providers.

hmm. and data transfer to america is fine?

deepseek made it possible to use the model offline as well. I dont see why we have to be hypocrits here.

edit: yes I know its about the app, and for the app the acusation is fair. but in general the news is not as hot as its written here.

appearently when people call they do no like to text. because they have no hands free or whatever

so voicemail is the better option than simply be pissed because i never called back.

nice take on this

If you call and I don’t pick up, leave a voicemail. If you don’t leave a voicemail, I assume it wasn’t important

this a thousand times. IDK when people decided to not use the voicemail anymore.

But to be honest I make it more radical and not even return all calls. because. priority 1 - call, not answered but with a voicemail

priority 2 - unanswered call but a message sent afterwards.

priority 3 - message only.

an unanswered call on my side and no further information is for me simply to forget about it.

das netz kommt in der tat mit dem ausbau von PV net klar. die netzregler laufen jz schon am limit. solang wir da nicht mehr ausbauen bringts sich nix mehr zu dezentralisieren.

it only enabled its operators access to encrypted communications...

what the heck? isnt this much worse than simple microphone access?

Wenn ich raten müsste eine Kombi aus:

den wirtschaftsmotor namens rene benko nicht vergessen der gerade stottert.

well. indeed the devil's in the detail.

But going with your story. Yes, you are right in general. But the human input is already there.

But you have to have human-made material to train the classifier, and if the classifier doesn’t improve, then the generator never does either.

AI can already understand what stripes are, and can draw the connection that a zebra is a horse without stripes. Therefore the human input is already given. Brute force learning will do the rest. Simply because time is irrelevant and computations occur at a much faster rate.

Therefore in the future I believe that AI will enhance itself. Because of the input it already got, which is sufficient to hone its skills.

While I know for now we are just talking about LLMs as blackboxes which are repetitive in generating output (no creativity). But the 2nd grader also has many skills which are sufficient to enlarge its knowledge. Not requiring everything taught by a human. in this sense.

I simply doubt this:

LLMs will get progressively less useful

Where will it get data about new programming languages or solutions to problems in new software?

On the other hand you are right. AI will not understand abstractions of something beyond its realm. But this does not mean it wont expedite in stuff that it can draw conclusions from.

And even in the case of new programming languages, I think a trained model will pick up the logic of the code - basically making use of its already learned pattern recognition skills. And probably at a faster pace than a human can understand a new programming language.

where?

https://forum.hyperion-entertainment.com/viewtopic.php?t=4733

does not work for me?

Well. I doubt that very much. Take as an analogy the success of the chess AI which was left training itself - compared to being trained...

Programmers as it turns out are very ‘eh, the code should explain itself to anyone with enough brains to look at it’ type of people

I cannot say how much I hate this.

even worse for old code where proper variable naming and underscores were forbidden. Impossible to get into someone else's head.

Puzzling Stack Exchange

this simply an aggregator?

People also blame ai, but if people are going to ai to ask the common already answered questions then… good!

exactly!

While I am indeed worried about the "wasted" energy (thats a whole other topic), thats pretty much why AI is good for.