Top computer scientists say the future of artificial intelligence is similar to that of Star Trek
Research fields con­tribut­ing to ShELL. Cred­it: Nature Machine Intel­li­gence (2024). DOI: 10.1038/s42256-024–00800‑2

Lead­ing com­put­er sci­en­tists from around the world have shared their vision for the future of arti­fi­cial intelligence—and it resem­bles the capa­bil­i­ties of Star Trek char­ac­ter “The Borg.”

Experts from the likes of Lough­bor­ough Uni­ver­si­ty, MIT, and Yale say we are set to see the emer­gence of “Col­lec­tive AI,” where numer­ous arti­fi­cial intel­li­gence units, each capa­ble of con­tin­u­ous­ly acquir­ing new knowl­edge and skills, form a net­work to share infor­ma­tion with each oth­er.

The researchers—who unveiled their vision in a per­spec­tive paper in Nature Machine Intel­li­gence—rec­og­nize the strik­ing sim­i­lar­i­ties between Col­lec­tive AI and many sci­ence fic­tion con­cepts. One exam­ple they cite is The Borg, cyber­net­ic organ­isms fea­tured in the Star Trek uni­verse, which oper­ate and share knowl­edge through a linked hive-mind.

How­ev­er, unlike many sci-fi nar­ra­tives, the com­put­er sci­en­tists envi­sion Col­lec­tive AI will lead to major pos­i­tive break­throughs across var­i­ous fields.

Lough­bor­ough Uni­ver­si­ty’s Dr. Andrea Soltog­gio, the research lead, explained, “Instant knowl­edge shar­ing across a col­lec­tive net­work of AI units capa­ble of con­tin­u­ous­ly learn­ing and adapt­ing to new data will enable rapid respons­es to nov­el sit­u­a­tions, chal­lenges, or threats.

“For exam­ple, in a cyber­se­cu­ri­ty set­ting if one AI unit iden­ti­fies a threat, it can quick­ly share knowl­edge and prompt a col­lec­tive response—much like how the human immune sys­tem pro­tects the body from out­side invaders.

“It could also lead to the devel­op­ment of dis­as­ter response robots that can quick­ly adapt to the con­di­tions they are dis­patched in, or per­son­al­ized med­ical agents that improve health out­comes by merg­ing cut­ting-edge med­ical knowl­edge with patient-spe­cif­ic infor­ma­tion.

“The poten­tial appli­ca­tions are vast and excit­ing.”

The researchers acknowl­edge there are risks asso­ci­at­ed with Col­lec­tive AI—such as the swift spread of poten­tial­ly uneth­i­cal or illic­it knowledge—but high­light a cru­cial safe­ty aspect of their vision: AI units main­tain their own objec­tives and inde­pen­dence from the col­lec­tive.

Dr. Soltog­gio says this would “result in a democ­ra­cy of AI agents, sig­nif­i­cant­ly reduc­ing the risks of an AI dom­i­na­tion by few large sys­tems.”

The com­put­er sci­en­tists arrived at the con­clu­sion that the future of AI lies in col­lec­tive intel­li­gence fol­low­ing an analy­sis of recent advance­ments in machine learn­ing.

Their research revealed glob­al efforts are con­cen­trat­ed on enabling life­long learn­ing (where an AI agent can extend its knowl­edge through­out its oper­a­tional lifes­pan) and devel­op­ing uni­ver­sal pro­to­cols and lan­guages that will allow AI sys­tems to share knowl­edge with each oth­er.

This dif­fers from cur­rent large AI mod­els, such as Chat­G­PT, which have lim­it­ed life­long learn­ing and knowl­edge-shar­ing capa­bil­i­ties. Such mod­els acquire most of their knowl­edge dur­ing ener­gy-intense train­ing ses­sions and are unable to con­tin­ue learn­ing.

“Recent research trends are extend­ing AI mod­els with the abil­i­ty to con­tin­u­ous­ly adapt once deployed, and make their knowl­edge reusable by oth­er mod­els, effec­tive­ly recy­cling knowl­edge to opti­mize learn­ing speed and ener­gy demands,” says Dr. Soltog­gio.

“We believe that the cur­rent dom­i­nat­ing large, expen­sive, non-share­able and non-life­long AI mod­els will not sur­vive in a future where sus­tain­able, evolv­ing, and shar­ing col­lec­tive of AI units are like­ly to emerge.”

He con­tin­ued, “Human knowl­edge has grown incre­men­tal­ly over mil­len­nia thanks to com­mu­ni­ca­tion and shar­ing.

“We believe sim­i­lar dynam­ics are like­ly to occur in future soci­eties of arti­fi­cial intel­li­gence units that will imple­ment demo­c­ra­t­ic and col­lab­o­rat­ing col­lec­tives.”

Vice-Chan­cel­lor and Pres­i­dent of Lough­bor­ough Uni­ver­si­ty, Pro­fes­sor Nick Jen­nings, is an inter­na­tion­al­ly-rec­og­nized author­i­ty in the areas of AI, autonomous sys­tems, cyber-secu­ri­ty and agent-based com­put­ing. He said, “I’m delight­ed to see Lough­bor­ough researchers lead­ing in this impor­tant area of AI research.

“This paper helps set the agen­da for the next wave of AI devel­op­ments, based upon mul­ti­ple, inter­act­ing agents. I look for­ward to see­ing this vision becom­ing a real­i­ty in the com­ing years.”

 




 

More information:A collective AI via lifelong learning and sharing at the edge, Nature Machine Intelligence (2024). DOI: 10.1038/s42256-024-00800-2. www.nature.com/articles/s42256-024-00800-2Provided byLoughborough UniversityCitation:Top computer scientists say the future of artificial intelligence is similar to that of Star Trek (2024, March 22)retrieved 3 April 2024from https://techxplore.com/news/2024-03-scientists-future-artificial-intelligence-similar.htmlThis document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, nopart may be reproduced without the written permission. The content is provided for information pur

poses only.