7UDGLQJ)DXUHV9LUWXDO0XVLFLDQVDQG0DFKLQH(WKLFV

1LFN&ROOLQV

Leonardo Music Journal, Volume 21, 2011, pp. 35-40 (Article)

3XEOLVKHGE\7KH0,73UHVV

For additional information about this article http://muse.jhu.edu/journals/lmj/summary/v021/21.collins01.html

Access provided by Durham University (6 Apr 2016 08:35 GMT) Trading Faures: Virtual Musicians and Machine Ethics

a b s t r a c t

Nick Collins Increased maturity in modeling human musicianship leads to many interesting artistic achieve- ments and challenges. This article takes the opportunity to reflect on future situations in which virtual musicians are traded like baseball cards, associated content-creator and autonomous musical agent rights, and the musical and moral conundrums that may result. Although many scenarios usical life may become very strange in- • Market forces promoting cer- presented here may seem far- M fetched with respect to the cur- deed as computer modelers continue to improve their vir- tain kinds of robot (for ex- tual emulations of human musicianship. Imagine the ability ample, highly compliant ones, rent level of artificial intelligence, it remains prudent and artisti- to share not only music but musicians, in a world of canned over more socially empowered cally stimulating to consider musical personalities, artist avatars and jobbing musical robots machines). them. Accepting basic human [1]. To paraphrase Aubrey de Grey [2], we might claim that curiosity and research teleol- there is a musician alive now whose commissioned simulacrum As the mention of debates about ogy, it is salutary to consider the more distant consequences will run millions of years in the future. Allowing for strong the immorality of computer games of our actions with respect to artificial intelligence (AI) scenarios [3], at some point, the might suggest, concerns over the in- aesthetics and ethics. systems may be autonomous enough to demand their own fluence of particular material and bank accounts, book their own gigs, chase or lead new musical censorship meant to control such trends, set their own riders. . . . materials has recurred throughout Much of this discourse may appear only an artificially in- recorded history (substitute videos, movies, books, plays, tri- telligent construction, a fanciful pipe organ dream. In one tones and philosophical oratory for computer games, etc.). line of argumentation, interactive music systems are but pro- Employment issues caused by technology recall the social up- grammers anticipating the likely decision options in real-time heaval of the Industrial Revolution, or more specifically for performance, and the machines do not themselves hold sub- rebellious literary robota, Karel Cˇ apek’s 1921 play “Rossum’s stantial independent musical insight. Yet machine musician- Universal Robots,” in which the “slave-worker” word robot was ship continues to advance, and machine learning techniques coined [7]. These themes are very familiar from science fic- may undermine many sureties here [4]. Overlapping general tion, and in turn Isaac Asimov’s laws of robotics have become a AI research, improvements in computer vision, speech rec- starting point for much academic discourse on machine ethics. ognition, humanoid embodiment and other aspects of social Most existing ethical debate in music technology centers on integration for machines can only complement the work of intellectual property issues [8], although employment issues computer music researchers acting to stretch our expectations have also arisen as recording technology has developed [9]. of musical autonomous systems. No musical robot has yet killed a human being, accidentally Anticipating such trends has an ethical dimension; the or otherwise. It will inevitably happen at some stage, whether young field of machine ethics (also sometimes termed robo- through an AI’s pure enthusiasm to play a distant and difficult ethics) considers such matters as the ethical consequences of note in defiance of the human body’s limits under Fitt’s law, access to convincing virtual simulations, and inchoate robot or as an aesthetic consequence of some future danger music rights [5]. For example, in a 2008 paper, Blay Whitby [6] con- movement led by robotic practitioners. In counterpoint, we siders such issues as: should not deny positives to new AI systems too, such as edu- cational and therapeutic applications, alongside new musical • The danger of antisocial tendencies developing through possibilities. Musical agents may support practice in social ex- private abuse of robotic assistants, a debate parallel to con- change within the “safe” domain of music. Systems may act as cerns over violent depictions in computer games intelligent tutors, to the extent of becoming musical familiars • The resources invested in complex robots (lost to other that grow up with their human companions. Undoubtedly, projects, wasted by mistreatment) musical AI has consequences throughout online and offline • The need for careful vetting of robots working with vulner- social activity. able people (such as the trend towards robotic caregivers The legal status of artificial intelligences will be negotiated for the elderly) in alliance with shifts in the capabilities and prominence of AI in society. Real-world cases are already arising in virtual prop- erty associated with virtual worlds with exchange mechanisms to real currencies [10]. Software agents are already used exten- sively in finance, and the act of delegating decision-making to Nick Collins (composer), University of Sussex, Department of Informatics, University of Sussex, Falmer, Brighton, BN1 9QJ, U.K. E-mail: . such an entity raises more problems the more independence See for supplemental files (such as audio and video) the agent takes on [11]. It is likely that legal issues around related to this issue of LMJ and accompanying CD. such systems will arise only gradually, with the continuing de-

©2011 ISAST LEONARDO MUSIC JOURNAL, Vol. 21, pp. 35–39, 2011 35 lay to really profound AI systems. Wood- software has also racked up massive sales, row Barfield outlines three cases fore- especially in Japan. In ’s glimpsing growing independence of case, we might ponder how to apportion automata: “the current status quo of credit between the original voice artist, property, the status of an indentured Saki Fujita, the anime designer Kei, the servant, and status and associated rights promoting company Crypton Future Me- of personhood” [12]. Agents may even- dia and the legions of fans who create tually act as free individuals, or at least content such as songs and videos [15]. in some commercial cases as employees, The fan mania surrounding these “idoru” where presently they tend to be treated [16] can be clearly seen in footage of a as tools (with the programmer totally 3D animation show in 2010: Canned accountable). Perhaps some interesting animations hold the crowd entirely in precedent cases will arise in the field of their sway, and inhuman effects, such as musical AI in due course. appearance and dissolution, only add to In preparing this article, I approached the theater of the show [17]. a number of robot-musician builders, This sort of tireless virtual musician interactive system designers and com- is becoming a staple of music technol- posers to solicit opinions. Most were ogy development, with the synthesis of conservative in their estimation of the the singing voice now a realistic com- strong AI scenario. Nevertheless, they mercial target. Many of the longer-term Fig. 1. Kei (anime artist), image of Hatsune often alluded to the inspirational role of implications were discussed with the Miku. (© , Inc. Used robot musicians, and in this spirit I pro- first commercial releases of the Voca- by permission. Sourced from .) around far distant musical personalities officially and unofficially creating vocal may enrich dialogue and adventure. So fonts based on famous singers, mixing let us not try to suppress all fantasy but and matching singer simulations across take a healthy attitude honoring human historical periods and incorporating vo- international jurisdictions, and in many curiosity. cal synthesis as but one element within a cases cartoon characters have better legal In reality, we already see phenomena complete algorithmic music system [18]. protection than people [22]. including intensive algorithmic model- No major international pop singer has In an interesting artistic response to ing of musical styles, autonomous musi- yet licensed his or her voice for Vocaloid, the issues of simulation, The Formant cal systems circulated by email as software although it is just a matter of time un- Brothers’ work Le Tombeau de Freddie patches, machine-listening analysis of au- til such a singer in early or late career (2009) virtuosically synthesizes a Freddie dio recordings for expressive attributes, makes this move (while this might seem Mercury-esque voice, which is made to and virtual anime pop stars. After first more likely to be in late career, an open sing “L’Internationale.” In the accompa- discussing past and contemporary prec- source early-career release might act as nying text manifesto, they characterize edents, we will extend our remit to more a fantastic promotional tool and get the this type of artwork as exploring “spec- speculative cases in the domain of science “original” lots of gigs). The reaction of ters without the dead,” providing the fiction and futurology, which may even- one of the artists who provided the first Japanese neologism roku-gaku (a deri- tually pose the real ethical conundrums wave of Vocaloid voices, Miriam Stockley, vation of “recorded music”). Masahiro and provide artistic game-changers. is highly interesting in the context of an Miwa clarified that the duo has had no ethical discussion of such technology: contact with Mercury’s estate but consid- ered their work fair comment: Virtual Musicians “At first I was quite horrified by the idea,” Already Exist Ms. Stockley said. “People tend to pay a If they claim legally their rights against lot of money to get my sound, and here The example of Gorillaz, a band fronted our synthesized voice, it would be very I am putting it on a font.” She changed exciting for us, because this means that by animated characters, is familiar; in re- her mind, she said, because “you can’t they recognize our synthesized voice as cent years, the rather boring musicians fight progress, no matter how strange ’s real voice. It shows behind the band have often appeared it sounds.” She also negotiated an un- various philosophical / esthetic / legal / disclosed percentage for each copy of onstage instead of their more interesting ethical problems concerning reality / vir- Miriam that sells [19]. tuality or human / machine. We would avatars. Ross Bagdasarian Sr.’s Alvin and be happy to discuss about them [23]. the Chipmunks had already anticipated Hatsune Miku is not an AI agent [20] much of this field in 1958, taking the fur- but more an avatar with a myriad of users. Their work fits well with the permissive ther step of an eponymous “chipmunk” Representational simulations of celebrity D¯ojin music scene in Japan also relevant vocal manipulation. Yet another ontolog- musicians have found their way into pop- to the Vocaloid virtual idols. There are ical progression has been committed by ular culture in another way recently; via earlier precedents in computer music, virtual divas such as Hatsune Miku (Fig. Guitar Hero. A controversy over a model for example Charles Dodge’s modeling 1), Kagamine Rin and Len, and Utatane of Kurt Cobain in Guitar Hero 5 saw the and repurposing of Caruso recordings in Piko, whose singing voices are driven by surviving members of Nirvana publicly “Any Resemblance Is Purely Coinciden- Yamaha’s Vocaloid software [13]. condemning the use of Cobain’s image, tal” (1980). Whilst manual sampling and Those who doubt the cultural impact with Harmonix claiming Cobain’s estate analysis-driven concatenative synthesis of such a contrivance should note that had sold the rights and Cobain’s widow’s may get us quite far in simulation, the the No. 1-charting album in Japan at the entering legal proceedings, with her ex- greatest generalizations come from au- end of May 2010 was sung by such virtual act role in the whole affair unresolved tomatically parametrizing spectral and idols [14]; associated content creation [21]. Image rights are not uniform across physical models, as with Dodge’s use of

36 Collins, Trading Faures Linear Predictive Coding for resynthesis, recordings (with some amount of psycho- virtual personalities for their personal or Vocaloid’s Spectral Modeling Synthe- logical and physiological modeling to fill use. Some regimes may censor virtual sis [24]. In interactive systems building, in the gaps) will only improve. musicians to their own political ortho- the potential to model living rather than doxies. Some automusicians may be held dead musicians is ethically pertinent. Bill at artificially high prices, proportional to Hsu notes of gradually building more ef- Future Scenarios perceived demand, while some unfortu- fective systems: From a survey of existing trends and nate historical re-enactments end up in practices, we stretch now futurewards. the bargain bins of future AI construct Suppose we learn to do this well enough Building really convincing virtual shops. Sore throats will no longer hinder so a John Russell android can be convinc- ing for 15–20 minutes. This is probably musical personalities (“musonalities”?) star singers, who will delegate their per- where I start worrying about whether I requires technology of undoubted mu- formance to indistinguishable simulacra should consult with John before I use sical impact. It would be hard to argue (but perhaps find they lose their fee to the system in public, make commercially that music would “stagnate” in the face the AI once they do so!). Really potent available recordings with it, etc. It’s not of such exciting technological facility, AIs may supplant human musicians [36] quite “identity theft,” but it would start to make me uncomfortable [25]. even if nostalgic re-enaction of historic through sheer virtuosity, reliability and performers becomes the “killer” (or “re- work ethic; they may prove exception- The line between influence and repro- incarnator”) app. Any worry about ex- ally popular with human audiences, if duction remains tense and uncertain. cessive loss of creativity, where humans they do not sublime into new musical The attraction of robot replace- defer their musical activities to agents, is realms beyond the ken of biological ments for musicians was outlined in the counter-balanced by the musical inven- hearing, where humans cannot follow. If late 1970s, both by Kraftwerk circa Die tiveness required to build such creations the strong AI position does prove war- Mensch-Maschine (1978) [26] and the in the first place and the mass human- ranted, musical AIs will be vulnerable to Human League’s proposed tour with agent creative ecosystem prompted by all the usual foibles of life; AI extortion, automatic orchestra [27]. Virtual anima- such capability. kidnapping, suicide, murder, plagiarism tions of humans have been increasingly In the digital realm, music is sensation- and more will enrich their biographies, accepted in culture [28], from the Max ally accessible, whether through browser alongside the trappings of fame, from Headroom character to the virtual news- or app-based streaming services, file shar- robot groupies to AI stalkers. caster Ananova. Avatar gigs in Second ing or download stores. Alongside fixed Algorithmic composition has often Life are commonplace, and in some ex- products, musical programs are also pointed to the merging of source data- perimental research projects seeking a readily available, and it is not hard to see bases and intersections of musical rules truer autonomy for the agents, artificial a further intensive trade in musical AIs [37]. Musical progression follows laws of musicians have been visually represented developing. Commerce and exchange memetics rather than genetics; aspects by virtual animated simulacra. Whilst we of pseudo-Mozarts raises more ethical of any musical period and style can be might still doubt the level of fluid musi- issues the more socially developed the brought into conflict or rapport with cal AI achieved, it is clear that the trend AI becomes (a new form of wearing liv- any other. Thus, we anticipate the cross- to autonomous and human-like virtual ery for a composer who dreamed of in- breeding of musical models, perhaps musicians is extremely well established, dependence). Although Mozart retains resolving old rivalries and tensions (Mo- both in computer-generated music and no musical copyright or image rights, a zart and Salieri), overturning prejudices graphics, and in explicit musical robotics potential eternity of musical slavery may (Wagner and klezmer) and spanning vast with acoustic synthesis. be enough to turn public opinion back time scales (Guido Spears and Britney Precedents in the use of an animated to “his” defense where branded Austrian d’Arezzo). Through sampling and com- computer character for musical interac- chocolates were not sufficient insult. Al- puter programming, such options are tion include representations of virtual though the “original” Mozart may never open to us now; the controversies of ran- musicians [29], a virtual head that ex- be a beneficiary, the endgame of the mu- sacking record collections hibits musical emotional cues [30], static sic AI project is to bring things to a head and creating musical programs have al- musical avatars automatically generated for “somebody”; the research goals are ready been playing out. Yet the ethical to reflect a user’s musical tastes [31] and most brilliantly met in exactly the case danger only increases in cases that involve a virtual conductor created for a human where the AI needs most rights. On the modeled personalities. What do musi- orchestra [32]. Robots have conducted other hand, whether future resources or cians think of collaborating with people orchestras as well (for instance, Honda’s political structures will effectively house whom they never met alive? Should the ASIMO in 2008 with the Detroit Sym- such debates remains to be witnessed. musical AI designer design out Wagner’s phony Orchestra) and have become an We can anticipate a host of social anti-Semitism as inappropriate or retain intensive center of new music research, changes, dangers and opportunities in it as a necessary and historically accurate often performing with humans [33]. trading musical models. Whether musi- flaw in his putative genius? Will willful Expressive modeling in alliance with ad- cians sign away new packages of rights historical re-enactments refuse to play vances in music information retrieval has to music software companies that virtu- with us after we have paid for collabora- a long-term goal of targeting the playing alize stars [35] or the companies trawl tion, but prove ourselves incapably slow styles of past musicians. There is a contin- history for their targets, clashes of inter- on the musical uptake? ual engineering drive to further the abili- est are bound to arise. Current tensions Whilst Robert Silverberg imagined a ties of transcription technology, whether between open source and closed source time machine bringing Pergolesi to the via Zenph’s semi-automatic restoration of will undoubtedly extend to virtual musi- future [38], he may as well have imagined old piano recordings, Celemony’s Melo- cian software. Export restrictions on AI a team of programmer-musicologists re- dyne products or cutting-edge academic software, and acts to prohibit tamper- constructing the past without breaking research [34]. Our ability to extract and ing, will not stamp out all illicit hacking. the laws of physics. Here we overlap with generalize musical personality from fixed Some collectors may hold back famous debates in the field of computational cre-

Collins, Trading Faures 37 ativity as to the extent to which domain- antees that no other relativistically ac- Agents,” Interacting with Computers 20 (2008) pp. 326– specific knowledge is ever isolated and cessible current time-frame is witnessing 333. There are also two interesting responses in that journal issue, from Alan Dix and Harold Thimbleby, whether the breathing, embodied hu- the running of a unique AI. A promoter worth reading to get contrasting viewpoints on the man being with bills to pay, novelists to might wish to advertise the only time you positive benefits of robot research and whether the date and children to feed is totally neces- will ever see Velvis take to the stage with field of machine ethics is “a seductive mix of fiction and reality.” See Harold Thimbleby, “Robot Ethics? sary to the whole emotional experience. Madonnatar. New meaning will gather Not Yet. A Reflection on Whitby’s ‘Sometimes It’s With lucrative commerce in virtual around cover bands, whilst if famous Hard to Be a Robot,’” Interacting with Computers 20 musicians comes lucrative lawsuits. Attri- musicians are produced cheaply enough (2008) pp. 338–341 (p. 338). bution of credit between programmers, (and pirated AIs may be commonplace) 7. For automation and morality, see p. 851 of Row- the original musicians and their estates, every child may grow up under musi- land Stout, “Twentieth-Century Moral Philosophy,” in Dermot Moran, ed., The Routledge Companion to their end clients and other tricky con- cal tutelage from a chorus of famous Twentieth Century Philosophy (New York: Routledge, tract deals will no doubt require an im- musicians [40]. 2008) pp. 851–882. For a literary treatment of au- tomation, see, for instance, Kurt Vonnegut’s Player mense medley of reanimated lawmakers Piano (New York: Charles Scribner’s Sons, 1952). gathered from across the ages. An array onclusions of transhuman law supermachines will C 8. For example: Lawrence Lessig, Free Culture: The The more of ourselves we commit to Nature and Future of Creativity (New York: Penguin broker the necessary rights across mul- Books, 2004); Paul D. Miller, ed., Sound Unbound: tiple real and virtual jurisdictions. Per- virtual musicians, the weirder and more Sampling Digital Music and Culture (Cambridge, MA: haps copyright will be perpetual within interesting musical life may become. The MIT Press, 2008); Mark A. McCutcheon, “Techno, Frankenstein and Copyright,” Popular Music 26, No. a few million years, or effectively lawless, John Oswaldo of Musical Robotics may 2, 259–280 (2007). or, most likely, will remain somewhere tour with a whole orchestra of appropri- ated musical personalities from through- 9. Unionized resistance includes campaigns against complexly in-between. Machines may sound on film, radio broadcast of recordings rather at some point stand up for their own IP out history, merging and mangling their than live musicians, sampling, and the introduction rights as dynamic creators, whilst the ex- personalities onstage in ways disturbing of MIDI. For a perspective from the early 1930s against “canned music,” see Mark Katz, Capturing isting big content companies will fight to many, alluring to some. Whether the Sound: How Technology Has Changed Music (Berkeley, to retain power as long as they can by AIs will revolt against their cruel fate, CA: Univ. of California Press, 2004) p. 67. or a concerned public of humans and denying that AIs have reached sufficient 10. F. Gregory Lastowka and Dan Hunter, “The Laws independence [39]. Imagine a war cued ahumans intervene, remains to be seen. of the Virtual Worlds,” California Law Review 92, No. by IP breaches: the “real” Bot Dylan ver- In a thought-provoking future, musical 1, 1–74 (2004). sus a pretender; more overtly, perhaps, systems with built-in moral nihilism may 11. Giovanni Sartor, “Cognitive Automata and the The Robotorious B.I.G. deleted by a rival compete with more conscientious AIs for Law: Electronic Contracting and the Intentionality of the No. 1 musical model of the week. Software Agents,” Artificial Intelligence Law 17 (2009) record company. . . . pp. 253–390. In a solar system–wide economy or beyond, where population figures run Acknowledgments 12. Woodrow Barfield, “Issues of Law or Software Agents within Virtual Environments,” Presence 14, to trillions, mass success may require Many thanks for stimulating discussion and feed- No. 6, 741–748 (2005). being in more than one place at once. back to Robert Rowe, Bill Hsu, Gil Weinberg, 13. Vocaloid was itself developed from research Musical AIs might be downloaded to lo- Seongah Shin, Godfried Willem Raes, Ajay Kapur, Akihiro Kubota, Masahiro Miwa and two anonymous in Barcelona at Universitat Pompeu Fabra. Jordi cal networks on different planets. This reviewers. Bonada and Xavier Serra, “Synthesis of the Singing would prove cheaper in distribution than Voice by Performance Sampling and Spectral Mod- els,” IEEE Signal Processing Magazine (March 2007). streaming fixed content and perhaps References and Notes conform better to data-sharing restraints 14. The album is Exit Tunes Presents Vocalogenesis Fea- 1. There are a profusion of terms in this field, such as turing Hatsune Miku, which reached No. 1 in the Ori- across a solar system. Fans may actively robots, virtual avatars, agents, AI constructs, machine con charts issued 31 May 2010. Retrieved 9 Dec 2010 assist the process of performance by vir- musicians, interactive music systems and the like. I ; tual idols by locally preparing physical shall use such terms relatively interchangeably in this . article, since the issues can apply across physical and substructs for AI musicians to download virtual space wherever artificial intelligence is found. 15. Aside from the Vocaloid software and fonts, there into. A compromise between local cus- are animation engines such as MikuMikuDance that 2. Randall Parker, “Aubrey de Grey: First Person to enable 3D model animation of idols. Japanese cul- tomization and minimum specification Live to 1000 Already Alive” (2004); Retrieved 27 ture is much more open to fan content creation than may lead to interesting conflicts and November 2010 from . Aubrey de Grey, “We Will Be derivative works to be built and have set up co-opting a musician mascot against that ber 2010 from . jp/>.There is also crossover from the Do¯jin music scene to the major labels, for example the Vocaloid promoting war, or a physical impairment 3. Stuart Russell and Peter Norvig, Artificial Intel- user collective Supercell, now signed to Sony Music ligence: A Modern Approach, 2nd Ed. (Upper Saddle cured or imposed. Entertainment Japan. River, NJ: Prentice Hall, 2003). Commercial models may be founded 16. William Gibson, Idoru (New York: G.P Putnam’s on paying for rent of virtual agent soft- 4. Nick Collins, “Reinforcement Learning for Live Musical Agents,” Proceedings of the International Com- Sons 1996). The novel imagines a marriage contract ware for personalized performance puter Music Conference, Belfast (2008). between a human singer and an AI virtual idol. situations. Whilst some companies may 5. Some moral philosophers would not admit this as a 17. “World Is Mine Live in HD,” retrieved 9 December abide by strict codes of use, others may new field, resting as it does on longstanding debates 2010, . see commercial advantage in allowing in ethics about the ultimate principles on which hu- 18. Bill Werde, “MUSIC; Could I Get That Song in the end-user greater freedom of choice man society operates. Machine ethicists often draw from debates in computer games or general tech- Elvis, Please?” The New York Times (23 November in how their agent can be customized. nological impact, for instance. For links to animal 2003), retrieved 9 December 2010 . The article discusses some may sell the uniqueness of performances, man? A Brief History of Humankind (Oxford, U.K.: of the arresting implications, including automatic both from the perspective of adaptation Oxford Univ. Press, 2004). popstar generation on a scale far past AutoTune tweaking. One producer is quoted as reacting, “It’s to local performance conditions (“Good 6. Blay Whitby, “Sometimes It’s Hard to Be a Robot: intriguing, this idea of ‘O.K., just give me all your evening, Charon!”) and based on guar- A Call for Action on the Ethics of Abusing Artificial vowels and all your consonants and I’ll see you later.’”

38 Collins, Trading Faures 19. Werde [18]. 29. R. Rowe and E. Singer, “Two Highly Integrated humans can do—thinking about music and playing Real-Time Music and Graphics Performance Sys- music in a way that we humans will never do.” Gil 20. The intelligence in Hatsune Miku is provided tems,” Proceedings of International Computer Music Con- Weinberg, personal communication, 30 November by the human content creator. As commentator so- ference, Thessaloniki, Greece, 1997; E. Singer et al., 2010. nicoliver put it in respect to a Hatsune Miku video “Improv: Interactive Improvisational Animation and on YouTube: “virtual body, virtual voice . . . now she Music,” International Symposium on Electronic Art, 1996; 37. David Cope’s Experiments in Musical Intelli- needs a virtual mind. . . .” posted 13 September 2010, R. Taylor, D. Torres and R. Boulanger, “Using Music gence come to mind, a project based around statis- retrieved 29 November 2010; the video has subsequently NIME, Vancouver, 2005. musical examples, for algorithmic composition of been deleted for copyright breaches associated with new works in a given style; if only the author had the uploader’s account. 30. M. Mancini, R. Bresin and C. Pelachaud, “A Vir- not destroyed the source databases. . . . Cope’s more tual Head Driven by Music Expressivity,” IEEE Trans- recent project Emily Howell explores the modeling 21. Sean Michaels, “Courtney Love to Sue over Kurt actions on Audio, Speech, and Language Processing 15, of creativity, one amongst many projects in the new Cobain Guitar Hero Appearance,” The Guardian No. 6, 1833–1841 (2007). field of computational creativity, a forwards-thinking (10 September 2009), retrieved 11 December 2010; branch of algorithmic composition. David Cope, ; Robin Murray, “Guitar Hero MA: MIT Press, 2005). Kurt Cobain Row” (2009), retrieved 11 December Content Description,” 5th Audio Mostly Conference: A 2010 from . 2010. Cocktail Party (London: VGSF 1989) pp. 152–170. 22. In the U.S.A, the “right of publicity” is the legal 32. D. Reidsma, A. Nijholt and P. Bos, “Temporal In- teraction between an Artificial Orchestra Conductor 39. Woodrow Barfield, “Intellectual Property Rights protection afforded. The U.K. is rather more loosely in Virtual Environments: Considering the Rights of legislated than the U.S., only covering image rights and Human Musicians,” Computers in Entertainment 6, Owners, Programmers, and Virtual Avatars,” Akron via such mechanisms as trademark law, with special No. 4, 1–22 (2008). Law Review, No. 39 (2006) pp. 649–700. conditions on use of likenesses in advertising but 33. See surveys in Ajay Kapur, “A History of Robotic greater freedom in satire, art and some commerce Musical Instruments,” Proceedings of the International 40. Ajay Kapur, personal communication, 23 No- when no attempt is made to claim the product is Computer Music Conference (ICMC), Barcelona, 2005; vember 2010: “Allowing globalization of master endorsed by the original (e.g. John Burns, “Celebrity Nick Collins, “Musical Robots and Listening Ma- knowledge on an instrument, rather than [where] Image Rights in Law” (2010), retrieved 20 Decem- chines,” in N. Collins and J. d’Escrivan, eds., The the selected few are fortunate enough to study one ber 2010 from ). As demonstration of the U.K.: Cambridge Univ. Press, 2007); Nick Collins, In- fair use of celebrity re-animation for satire, the mil- troduction to Computer Music (Chichester, U.K.: Wiley, lennial U.K. Channel 4 comedy animation “House 2009). of Rock” explored afterlife representations of dead Manuscript received 1 January 2011. rock stars. 34. Anssi Klapuri and Manuel Davy, eds., Signal Processing Methods for Music Transcription (New York: 23. Masahiro Miwa, personal communication, 10 Nick Collins is an experienced composer, per- Springer, 2006). December 2010. former and researcher, with interests including 35. Imagine access to musical agent technology being machine listening, interactive and generative 24. In the context of Vocaloid models, automatic cus- preconditioned on your signing away your own rights The tomization to the vocal mannerisms of a guide vocal- music, and audiovisuals. He co-edited to your unique musical development; that is, to be Cambridge Companion to Electronic ist is discussed in Tomoyasu Nakano and Masataka tutored by advanced virtual agents, you must accept Goto, “VocaListener: A Singing-to-Singing Synthesis that your mature musical self is itself fully open to Music (Cambridge University Press, 2007) System Based on Iterative Parameter Estimation,” modeling. and The SuperCollider Book (MIT Press, Proceedings of the 6th Sound and Music Computing Con- 2011) and wrote The Introduction to ference (2009) pp. 343–348. 36. The ethical debate here will be important. “I would see it as ethically problematic if robots that Computer Music (Wiley, 2009). iPhone 25. Bill Hsu, personal communication, 12 December attempt to play like humans will start to replace hu- apps include RISCy, Concat, BBCut and 2010. man musicians in concerts today . . . robots are still PhotoNoise for iPad. Notable concerts include not as good as humans, so the outcome will be in- 26. Interestingly, the German Metropolis influence live coding in a vineyard in Corfu, falling off ferior. But more importantly, I think it is a waste of also arises in Ultravox’s 1977 track “I Want to Be a research resources to attempt to make robots who a piano stool in Sydney, and singing the 100 Machine.” play like humans. It may be interesting in terms of meter in Brighton. Sometimes, he writes in the 27. Simon Reynolds, Rip It Up and Start Again (Lon- theoretical research, i.e. to model how we humans third person about himself, but is trying to give don: Faber and Faber Limited, 2005). play music to enrich our knowledge about the cogni- it up. Further details, including publications, tive and physical aspects of this unique phenomena. 28. Sidney Eve Matrix, Cyberpop: Digital Lifestyles and But it is so much more interesting to try to develop music, code and more, are available from Commodity Culture (New York: Routledge, 2006). robotic musicians that push the envelope of what .

Collins, Trading Faures 39 announcement

Artists and Scientists on the Cultural Context of Climate Change

Leonardo explores the ways in which artists and scientists are addressing climate change. As contemporary culture grapples with this critical global issue, Leonardo has documented cross-disciplinary explorations by artists, scientists and engineers, working alone or in teams, addressing themes related to global warming and climate change.

Partial list of Leonardo articles and projects concerned with global warming, climate change and related issues:

Chris Welsby, “Technology, Nature, Soft- George Gessert, “Gathered from Coinci- Andrea Polli, “Heat and the Heartbeat of ware and Networks: Materializing the dence: Reflections on Art in a Time of the City: Sonifying Data Describing Cli- Post-Romantic Landscape,” Leonardo 44, Global Warming,” Leonardo 40, No. 3, mate Change,” Leonardo Music Journal 16 No. 2 (2011). (2007). (2006). Andrea Polli and Joe Gilmore, “N. Special Section: Environment 2.0, Guest Editor Julien Knebusch, “Art & Climate Change,” April 16, 2006,” LMJ16 CD Contributor's Drew Hemment. Authors include Carlo Web project of the French Leonardo Note, Leonardo Music Journal 16 (2006). Buontempo, Alfie Dennen, Yara El- group Leonardo/Olats (l'Observatoire Sherbini, Rebecca Ellis, Drew Hemment, Leonardo pour les Arts et les Techno- Janine Randerson, “Between Reason and Christian Nold, John Tweddle and Brian Sciences), . Climate Change,” Leonardo 40, No. 5 W. Paul Adderley and Michael Young, (2007). Julien Knebusch, “The Perception of Cli- “Ground-breaking: Scientific and Sonic Per- Ruth Wallen, “Of Story and Place: Com- mate Change,” Leonardo 40, No. 2 (2007). ceptions of Environmental Change in the municating Ecological Principles through African Sahel,” Leonardo 42, No. 5 (2009). Art,” Leonardo 36, No. 3, (2003). Andrea Polli, “Atmospherics/Weather David D. Dunn and James P. Crutchfield, Works: A Spatialized Meteorological Data Angelo Stagno and Andrea van der “Entomogenic Climate Change: Insect Sonification Project,”Leonardo 38, No. 1, Straeten, “0-24 Licht: A Project Combining Bioacoustics and Future Forest Ecology,” (2005). Art and Applied Research,” Leonardo 40, Leonardo 42, No. 4 (2009). No. 5 (2007).