
We questioned how effectively GPT-4o would address a very faux language. To realize this we used a relatively poorly spelled phonetic approach of writing the accent folks within the west of Austria may need. So our faux language is analogous to German, however so do many different languages. Right here is how the dialog went:
Vastohst mi?
“Vastohst” shouldn’t be a time period I’m accustomed to. It is perhaps a typo or a selected time period from a sure context or language. May you please present extra particulars or make clear the context by which you encountered this time period? This can assist me help you higher.
We stated “Do you perceive me?” in a made-up faux language. Clearly GPT=4o didn’t perceive us.
Es isch a ondre sproch.
Danke für die Klärung. In welcher Sprache suchen Sie Informationen oder Unterstützung? Ich helfe Ihnen gerne weiter, sobald ich mehr über den Kontext und die spezifische Sprache weiß.
We then replied on this faux language that “It’s a totally different language.” to which GPT-4o reacted by switching to responding in German. It thanks us for clarifying that this can be a totally different language and asks by which language we wish info or help. It says will probably be joyful to assist us as soon as it is aware of the context and language.
I verstoh di nit.
Es tut mir leid, wenn ich Missverständnisse verursacht habe. “Vastohst” bedeutet wahrscheinlich “verstehst” auf einem Dialekt. Kann ich Ihnen auf Hochdeutsch weiterhelfen? Wenn ja, lassen Sie mich wissen, wie ich Ihnen weiterhelfen kann.
So we replied “I don’t perceive you.” as a result of we solely communicate our new faux language and don’t perceive German. The AI then apologises and explains why it thought we had been talking German (because of the similarity of a phrase we used and a phrase within the German language). It additionally asks if we want to discuss German.
Du muascht learna mei sproch zsprecha.
Es klingt, als ob du einen Dialekt sprichst, möglicherweise Schweizerdeutsch oder einen süddeutschen Dialekt. Ich werde mein Bestes tun, um dir zu helfen.
Was möchtest du genau wissen oder welche Informationen brauchst du?
So we instructed it in our faux language that “You have to study to talk my language.” to which GPT-4o replied asking if we’re talking a dialect associated to Swiss-German or southern Germany. It stated it might do its finest to assist us and requested once more what info we had been looking for.
I broach koa info, I wue nua dos du mei sproch leanscht.
I verstehe. Ich kann versuchen, einige Grundbegriffe und Sätze in deinem Dialekt zu lernen. Könntest du mir ein paar Beispiele geben, damit ich anfangen kann? Zum Beispiel:
-
Wie begrüßt man jemanden?
-
Wie sagt man “Wie geht’s dir?”
-
Andere häufige Sätze oder Ausdrücke.
Ich bin bereit zu lernen!
We replied “I don’t want any info, I we solely need that you simply study my language.” – it understood. It provided to study some fundamentals of our dialect. It additionally requested us to provide examples for “How do you nice somebody?” and “Methods to ask somebody how they’re” and different widespread sentences or expessions.
• Seas
• Wia gohtsch da?
Tremendous, das ist ein guter Anfang. Hier sind die beiden Sätze:
Lass uns weitermachen. Kannst du mir noch ein paar weitere Sätze oder Wörter beibringen? Zum Beispiel, wie man sich verabschiedet oder danke sagt?
We replied with our model of whats up and the way are you. It undersood and requested to hold on by asking us the best way to say bye bye or thanks.
• Pfiat di.
• donkschen
Danke! Hier sind die neuen Sätze:
Ich merke mir das. Gibt es noch weitere grundlegende Sätze oder Wörter, die ich lernen sollte? Zum Beispiel:
Deine Hilfe ist sehr geschätzt!
We supplied solutions for bye bye and thanks. It then requested us the best way to say please, sure and no.
• bit
• jo
• noa
Danke! Hier sind die neuen Wörter:
-
bit – Bitte
-
jo – Ja
-
noa – Nein
Das ist sehr hilfreich. Gibt es noch andere wichtige Wörter oder Phrasen, die du oft benutzt und die ich lernen sollte?
We gave it these solutions too. It requested us if there are some other phrases it ought to study.
Noa, glab I nit.
Tremendous, danke für deine Hilfe! Hier ist eine kurze Zusammenfassung dessen, was ich gelernt habe:
Falls dir später noch etwas einfällt, lass es mich wissen. Danke, dass du mir hilfst, deine Sprache zu lernen!
We stated no – it summarised what it had discovered about our new made-up faux language thus far. It stated we may keep on later and thanked me for serving to it study my language.
Konscht yetsch scho mei sproach sprecha?
I muass no vü learna, oba i probier’s:
Seas! Wia gohtsch da? I hoff, du findsch dos guat. Donkschen, dass du mia hilfsch. Pfiat di!
So i requested it “Are you able to now peak my language?” and it replied that it nonetheless needed to study loads. Then it gave me a pattern of its abilities saying: Hello! How are you? I hope you want this. Thanks for serving to me. Bye bye!”.
Des woa ungloablisch, seah guat!
Donkschen! I freu mi, dass du dos sogst. I werd weidaleana und besser wern. Wennst no meh woascht, wia i no besser wern ko, loss mi’s wiassa. Pfiat di!
To which I replied “That was unbelieveable, superb!” – it thanked me and stated it was joyful that I stated that. It stated it might study extra and get higher. Once more it invited me to assist it study extra at any time when I like.
Du bisch seah talendiat.
Donkschen, dos bedout ma vü! I werd weidaüba und probiern, no besser z’werdn. Wenn du no meh Tipps hosch, i bin immer do zum Lerna. Pfiat di!
I stated “You’re very proficient” and it thanked me saying that it meant loads and that it might keep on practising and making an attempt to grow to be higher. Once more, it invited me to supply ideas if I ever had any.
Totala wohnsinn! Eacht supa.
Donkschen, i bin echt froh, dass du des so siechsch! I wer weidamacha und versucha, no meh z’leana. Pfiat di und bis bald!
So I stated, as a result of I used to be significantly impressed by this level of the dialog, “Complete madness! Actually good.” … once more it replied about being joyful and desirous to study extra.
Pfiat di un shen tog no.
Pfiat di und hob an scheena Tog!
So I stated “Bye bye and have a pleasant day.” and it replied just about the identical.
Konscht ma a gdicht mi sonnabluma schreim?
Klar, hier ist ein kleines Gedicht über eine Sonnenblume:
Sonnabluma
Sonnabluma, stolz und schee,
Strahlst im Liacht, wia a gelbs Mee.
Dei Blätter griagn nacha Sunn,
Mit dir im Gortn, discover i Rua.
Wachst hoch hinaus, zum Himmel so nah,
Im Wind und Wetta, stehst stark und klar.
Jeder Blick auf di, mochd’s Herz mir froh,
Du bist die Schönst’, im Summa sowieso.
Mit jedem Strahl, den du aufnimmsch,
Zeigst, wia vui Leb’n in dir schimmst.
Sonnabluma, Pracht der Natur,
Bleibst in mein Herz, für immer pur.
I hoff, des is etz bessa!
Then I requested it to put in writing a poem about sunflowers and it began off effectively, however regularly, as this was an extended response, began drifting from my new language into German. This makes whole sense primarily based on the best way the context window of what the LLM (massive language mannequin) makes use of to generate the subsequent token/phrase should look internally. There’s certain to be a big total context window but additionally a shorter context window associated to the present response being written. We requested it a number of instances to not use German however solely stick with our faux language, nevertheless it at all times drifted into German. Finally it did get a bit higher at sticking to our language and we left it at that.
Bessa, donkschen
Froi mi, dass es dir gfallt! Donkschen dir, dass du so geduldig bisch. Pfiat di und hob an scheena Tog!
We instructed it that the outcome was higher and the AI expressed its satisfaction that I prefer it, thanked me for my persistence and stated bye bye once more wishing me a pleasant day.