Ar­ti­fi­cial in­tel­li­gence driven by Softbank’s trea­sure chest

Re­gard­less of how ar­ti­fi­cial in­tel­li­gence (AI) is de­fined, there is lit­tle doubt that this re­source can be of great value, es­pe­cially in big data ap­pli­ca­tions.

The Malta Independent on Sunday - - BUSINESS & FINANCE - Ge­orge M. Man­gion

Un­doubt­edly, AI is fast be­com­ing a ma­jor tech­no­log­i­cal tool for pre­scrip­tive an­a­lyt­ics, the step be­yond pre­dic­tive an­a­lyt­ics that helps us de­ter­mine how to im­ple­ment and/or op­ti­mize op­ti­mal de­ci­sions. In busi­ness ap­pli­ca­tions, it can as­sess fu­ture risks, quan­tify prob­a­bil­i­ties and in so do­ing, give us in­sights how to im­prove mar­ket pen­e­tra­tion, cus­tomer sat­is­fac­tion, se­cu­rity anal­y­sis, trade ex­e­cu­tion, fraud de­tec­tion and preven­tion, while prov­ing in­dis­pens­able in land and air traf­fic con­trol, na­tional se­cu­rity and de­fence. There are also a host of health­care ap­pli­ca­tions such as pa­tient-spe­cific treat­ments for dis­eases and ill­nesses.

Re­cently, the pop­u­lar con­cept of “Sin­gu­lar­ity” was per­ceived by com­puter sci­en­tists. The idea was for­mally coined in 1993 by Ver­nor Vinge, a sci­en­tist and science fic­tion writer, who posited that ac­cel­er­at­ing tech­no­log­i­cal change would in­evitably lead to ma­chine in­tel­li­gence that would match and then sur­pass hu­man in­tel­li­gence. In his orig­i­nal es­say, Dr Vinge sug­gested that the point in time at which ma­chines at­tained su­per­hu­man in­tel­li­gence would hap­pen some­time be­tween 2005 and 2030. The no­tion of the “Sin­gu­lar­ity” is pred­i­cated on Moore’s Law, the 1965 ob­ser­va­tion by In­tel co­founder Gor­don Moore, that the num­ber of tran­sis­tors that can be etched onto a sliver of sil­i­con dou­bles at roughly two year in­ter­vals. This has fos­tered the no­tion of ex­po­nen­tial change, in which tech­nol­ogy ad­vances slowly at first and then with in­creas­ing ra­pid­ity with each suc­ceed­ing tech­no­log­i­cal gen­er­a­tion.

To men­tion a few ex­am­ples of the rapid progress made by re­search and de­vel­op­ment co­funded by multi-na­tional firms, one can start by men­tion­ing Google. The gi­ant search en­gine firm is a pioneer in the field of ar­ti­fi­cial in­tel­li­gence, de­vel­op­ing self-driv­ing au­to­mo­biles, smart­phone as­sis­tants and other ex­am­ples of ma­chine learn­ing. Equally omi­nous was the pre­dic­tion four years ago by Prof. Hawk­ing who said the prim­i­tive forms of ar­ti­fi­cial in­tel­li­gence de­vel­oped so far have al­ready proved very use­ful, but he fears the con­se­quences of cre­at­ing some­thing that can match or sur­pass hu­mans. “Hu­mans, who are lim­ited by slow bi­o­log­i­cal evo­lu­tion, couldn’t com­pete, and would be su­per­seded.” Oth­ers think this warn­ing is too pes­simistic and ar­gue that we are a long way from hav­ing the com­put­ing power or the abil­ity to de­velop the al­go­rithms needed to achieve full ar­ti­fi­cial in­tel­li­gence. It will not hit us yet for a num­ber of decades.

On the con­trary, there is al­ways the fear of what the fu­ture may bring since we can­not quite know what will hap­pen if a ma­chine ex­ceeds our own in­tel­li­gence. We can­not know if we will be in­fin­itely helped by it, or ig­nored by it or, con­ceiv­ably, de­stroyed by it. When it comes to us­ing ar­ti­fi­cial in­tel­li­gence to power com­plex ro­bot­ics, one can­not ig­nore the worst fears of prom­i­nent tech­nol­o­gists and sci­en­tists like Elon Musk, Stephen Hawk­ing and Bill Gates, who have all voiced alarm over the pos­si­ble emer­gence of self­aware ma­chines which un­less har­nessed, may well be out to do harm to the hu­man race. Mr Musk of Tesla fame said: “If I had to guess what our big­gest ex­is­ten­tial threat is, it’s prob­a­bly AI”. In a cau­tion­ary mood of ad­mo­ni­tion, he has said that ar­ti­fi­cial in­tel­li­gence would “sum­mon the de­mon”.

One may ask who is fund­ing such ex­pen­sive re­search. The an­swer is a co­hort of ven­ture cap­i­tal­ists who are con­stantly poised to look out for tal­ented peo­ple in their on­go­ing re­cruit­ing outreach. It is not un­com­mon for re­search firms to seek top-notch uni­ver­sity grad­u­ates who show lead­er­ship po­ten­tial. Mr Masayoshi Son (see pic­ture) is a Ja­panese in­vestor who cre- ated SoftBank which he wants to mimic a “vir­tual Sil­i­con Val­ley”, mean­ing a plat­form on which uni­corns (start-ups that turned into a bil­lion dol­lar marvel) can of­fer each other con­tacts and ad­vice, buy goods and ser­vices from each other, and even join forces. Mr Son, who founded SoftBank in the 1980s, has grand vi­sions of what tech­no­log­i­cal ad­vance­ments the fu­ture holds. SoftBank’s sub­sidiaries are push­ing the fron­tiers of tech­nol­ogy in ar­eas such as the “In­ter­net of Things”, ar­ti­fi­cial in­tel­li­gence and deep learn­ing. It hatched the unique “Vi­sion Fund” with a $100 bil­lion war chest look­ing to in­vest in star­tups with op­er­a­tional ex­pe­ri­ence, and tech­ni­cal back­ground.

For this pur­pose, the Vi­sion Fund is ag­gres­sively com­pet­ing with tra­di­tional tech­nol­ogy in­vestors in Sil­i­con Val­ley in a no-holds-barred fight for tal­ent. Mr Son be­lieves he has a unique abil­ity to pre­dict fu­ture tech­nol­ogy trends, and states he is ready for the gam­ble. SoftBank is syn­ony­mous with its charis­matic founder that is re­shap­ing global tech with its colos­sal trea­sure box. It is shak­ing up the cosy world of Sil­i­con Val­ley ven­ture cap­i­tal. The gar­gan­tuan fund lures start-ups to cash out from the clutches of Google, Face­book and Ama­zon – hav­ing its mas­sive cheque­book it gives en­trepreneurs a bet­ter shot at com­pet­ing with the ti­tans. The fund wants to per­form a sim­i­lar func­tion in China, where nearly half of all uni­corns are now backed by one of the coun­try’s four tech gi­ants – Baidu, Alibaba, Ten­cent or JD.com.

In pass­ing, one may feel that the dis­rup­tive path of new tech­nol­ogy can­not find a bet­ter cham­pion than SoftBank with its ag­gres­sive in­vest­ment ap­petite to nur­ture new ven­tures. Read­ers ap­pre­ci­ate that this dis­rup­tive tech­nol­ogy has a be­nign pur­pose and is help­ing to link var­i­ous civ­i­liza­tions, im­prove crop yields and speed up the progress in com­plex hu­man Genome clas­si­fi­ca­tion. De­liv­ery drones, both wheeled and air­borne may in the near fu­ture com­pete with couri­ers while su­per­mar­ket ro­bots silently stack food items on shelves and move mer­chan­dise in ware­houses. Ar­ti­fi­cial in­tel­li­gence in ma­chines can even repli­cate hu­man judg­ments pre­vi­ously con­sid­ered to be too com­plex. Imag­ine how in the next decade, there will be ro­bots which are ef­fi­cient and de­void of emo­tions qui­etly su­per­vis­ing hun­dreds of com­plex fac­tory op­er­a­tions. In con­clu­sion, as if by magic, the Vi­sion Fund will help hu­man­ity get ma­chines able to de­velop com­plex al­go­rithms that ‘learn’ from past ex­am­ples – only then can Mr Son hap­pily claim that he has ful­filled his dream.

Newspapers in English

Newspapers from Malta

© PressReader. All rights reserved.