In diesem Vortrag am MIT vom 22.3.2023 stellt Sebastién Bubeck die Ergebnisse der Tests vor, die sein Team mit GPT-4 durchgeführt hat. Sehr interessant, auch wenn Musik nur ganz am Ende kurz erwähnt wird.
Es gibt jedoch einen sozusagen begleitenden Artikel auf arXiv, in dem auf Seite 19/20 auch auf das Thema Musik eingegangen wird. Dort heißt es:
2.3 Music
The data on which the model was trained also contains musical information encoded as ABC notation. This is a system that uses letters, numbers and symbols to represent musical pitches, durations, chords and other elements in a compact and readable way. We are interested in exploring how well the model has acquired musical skills from this exposure, such as composing new melodies, transforming existing ones, and understanding musical patterns and structures.
When instructed to generate a short tune (Figure 2.9), and the model was able to produce valid ABC notation. The tune had a clear structure, the time signature was consistent between bars and the notes followed increasing and decreasing patterns. The tune also used a consistent set of notes within the melody, and the rhythm had a repetitive pattern. However, the model did not seem to obtain the skill of understanding harmony. In fact, consecutive notes in the generated tunes are almost always adjacent to each other (namely, the note following C will almost typically be either B or D), and testing on 10 generated tunes, we were not able to extract any clear chords or arpeggios.
We then asked the model to describe the tune in musical terms. It was able to successfully give a technical description of the structure in terms of repetitions, the rising or descending parts of the melody and to some extent the rhythm. However, it seems that the descriptions of the harmony and chords are not consistent with the notes (in fact, it refers to sequences of adjacent notes, which do not form valid chords, as arpeggios).
We then asked the model to manipulate the melody in two ways. First, we instructed to change a certain rising sequence to a descending one, which it did successfully. Then we asked the model to convert the tune to a duet adding a bass voice. The model successfully extends the ABC notation with a second staff which has compatible rhythm and is played on a lower octave, however there is a lack of harmony between the two voices.
In summary, the model was able to produce valid tunes in ABC notation and, to some extent, explain and manipulate their structure. However, we could not get the model to produce any nontrivial form of harmony. It should be noted that ABC notation is not a very widely used format, and in fact the model was not able to produce even the most well-known tunes in ABC notation (such as Ode to Joy, Für Elise or Greensleeves, all of which are abundant online in that format), nor was it able to recognize these tunes.
HUMAN: Can you compose a short tune (say four to eight bars) using ABC notation?
GPT-4: There are many possible ways to compose a short tune in abc notation, but here is one example:
X:1
T:Example Tune
M:4/4
L:1/8
K:C
|:G2AB c2BA | G2AB c2c2 | edcB c2BA | G4-G2z2 :|
|:c2de f2ed | c2de f2f2 | g2fe d2cB | c4-c2z2 :|
This tune has two four-bar phrases, each repeated once, forming an A part and a B part. The key is C
major.HUMAN: Can you describe the tune in musical terms?
GPT-4: Here is one possible description:
The tune starts with a rising arpeggio of the tonic chord (C major), followed by a descending scale that
returns to the tonic. This pattern is repeated, but the second time the scale continues to the higher
tonic note. The first phrase ends with a long tonic note.
The second phrase begins with a similar pattern, but using the subdominant chord (F major) instead of
the tonic. The second time, the scale continues to the higher subdominant note. The phrase ends with a
stepwise descent to the tonic, followed by a long tonic note.
The A part and the B part have a similar structure, but use different chords and ranges. The tune has a
simple and catchy melody, with a clear sense of tonality and rhythm.HUMAN: I like part A, but maybe part B is just a bit too similar to part A, making the song repetitive. How
about beginning part B with a descending arpeggio instead of a rising one?GPT-4: Sure, that would create some contrast and variation. Here is one possible way to do that:
X:1
T:Example Tune
M:4/4
L:1/8
K:C
|:G2AB c2BA | G2AB c2c2 | edcB c2BA | G4-G2z2 :|
|:f2ed c2BA | f2ed c2c2 | g2fe d2cB | c4-c2z2 :|HUMAN: That’s much better. Now how about making this into a duet, can you add a second staff for this with a
bass accompanying the melody?GPT-4: One possible way to write a second staff for this tune is to use root notes and fifths of the chords, as
well as some passing notes, to create a simple accompaniment. Here is an example in abc notation, with
the second hand bass on the lower staff:
X:1
T:Example Tune
M:4/4
L:1/8
K:C
V:1
|:G2AB c2BA | G2AB c2c2 | edcB c2BA | G4-G2z2 :|
|:f2ed c2BA | f2ed c2c2 | g2fe d2cB | c4-c2z2 :|
V:2 clef=bass
|:C,2C,2 G,2G,2 | C,2C,2 G,2G,2 | F,2F,2 E,2E,2 | C,4-C,2z2 :|
|:F,2F,2 C,2C,2 | F,2F,2 C,2C,2 | G,2G,2 F,2F,2 | E,2C,2 G,2C,2 :|
Figure 2.9: Generating and modifying a tune in ABC notation.
--
"I spent a lot of money on booze, birds and fast cars. The rest I just squandered." - George Best --- Dienstags und donnerstags, ab 20 Uhr, samstags ab 20.30 Uhr: Radio StoneFM