Chati yaGartner yakafanana neshow yefashoni kune avo vari muhunyanzvi. Inokuudza pachine nguva kuti api anopisa mabuzzwords mwaka uno uye zvauchanzwa pamisangano yese iri kuuya.

Tasarudza zviri kuseri kwemazwi akanaka ari muchati ino kuti iwe ugone kutaurawo mutauro wacho.

Kutanga, mazwi mashoma pamusoro peiyi girafu. Gore rega rega muna Nyamavhuvhu, iyo yekubvunza agency Gartner inoburitsa chirevo - Gartner Hype Curve. MuchiRussia, iyi ndiyo "hype curve", kana, zviri nyore, hype. Makore makumi matatu apfuura, vaimbi vechikwata chePublic Enemy vakaimba kuti: "Musatenda hype". Kunyangwe kuzvitenda kana kusazvitenda inyaya yako pachako, asi zvakakosha kuti uzive aya mazwi akakosha kana ukashanda muindasitiri yehunyanzvi uye uchida kuziva mafambiro epasirese.

Iyi igirafu yezvinotarisirwa neveruzhinji zveimwe tekinoroji. Sekureva kwaGartner, zvine mutsindo tekinoroji inogara ichipfuura nematanho mashanu: kutangwa kwetekinoroji, kukwirira kwetarisiro yakakwira, mupata wekuodzwa mwoyo, kutsetseka kwechiedza, bani rekubudirira. Asi zvinoitikawo kuti inonyura mu "mupata wekuodzwa mwoyo" - iwe unogona nyore kuyeuka mienzaniso iwe pachako, tora bitcoins imwechete: pakutanga kusvika pakakwirira se "mari yeramangwana", vakakurumidza kudonhedzera pasi apo kukanganisa kwekombiyuta kwakava pachena, kunyanya zvirambidzo pahuwandu hwekutengeserana uye huwandu hwemagetsi hunodiwa kubudisa bitcoins (iyo matambudziko ezvakatipoteredza kare). Uye zvechokwadi, hatifanire kukanganwa kuti Gartner girafu ingori fungidziro: pano, semuenzaniso, iwe unogona kuverenga yakadzama. chinyorwa, uko kufanotaura kunoshamisa kusina kuzadzika kunoongororwa.

Saka, ngatimhanyei nechati itsva yeGartner. Tekinoroji dzakakamurwa kuita 5 hombe thematic mapoka:

1. Yepamberi AI uye Analytics
2. Postclassical Compute uye Comms
3. Sensing uye Mobility
4. Augmented Human
5. Digital Ecosystems

1. Advanced AI uye Analytics

Makore gumi apfuura akaona awa yakanakisa yekudzidza kwakadzama. Manetiweki aya anonyatso shanda kune awo akasiyana emabasa. Muna 10, Yann LeCun, Geoffrey Hinton naYoshua Bengio vakagamuchira Mubairo weTuring wezvavakawana mavari - mubairo une mukurumbira, wakaenzana neNobel Prize musainzi yekombuta. Saka, maitiro makuru munzvimbo ino, ayo akarongwa pane girafu:

1.1 Kutamisa Kudzidza

Iwe haudzidzise neural network kubva kutanga, iwe unotora yakatodzidziswa woipa imwe chinangwa. Dzimwe nguva izvi zvinoda kudzokorora chikamu chetiweki, asi kwete network yese, iyo inokurumidza kukurumidza. Semuenzaniso, kutora yakagadzirira-yakagadzirwa neural network ResNet50, yakadzidziswa paImageNet1000 dataset, unowana algorithm inokwanisa kurongedza zvinhu zvakawanda zvakasiyana nemufananidzo padanho rakadzama kwazvo (makirasi 1000 anoenderana nemaitiro anogadzirwa nezvikamu makumi mashanu zveneural network). Asi haufanire kudzidzisa iyi network yese, izvo zvingatora mwedzi.

В online course Samsung's "Neural Networks uye Computer Vision," semuenzaniso, iri mukupedzisira Kaggle task nekuiswa kwemaplateti kuita akachena uye ane tsvina, nzira inoratidzwa kuti mumaminetsi mashanu inokupa yakadzika neural network inokwanisa kusiyanisa mahwendefa ane tsvina kubva kune akachena, akavakirwa pazvivakwa zvinotsanangurwa pamusoro. Iyo yekutanga network yaisaziva kuti mahwendefa aive api zvachose, yakangodzidza kusiyanisa shiri nembwa (ona ImageNet).

Source: online course Samsung "Neural Networks uye Computer Vision"

Kuendesa Kudzidza kunoda ruzivo rwekuti ndedzipi nzira dzinoshanda, ndedzipi dzakagadzirirwa-dzakagadzirwa zvivakwa zvekutanga. Kazhinji, izvi zvinomhanyisa zvakanyanya kubuda kweanoshanda mashandisirwo ekudzidza muchina.

1.2. Generative Adversarial Networks (GAN)

Izvi ndezvezviitiko izvo pazvinenge zvakaoma kuti isu tigadzire chinangwa chekudzidza. Iyo iri padyo nebasa iri kune hupenyu chaihwo, iyo yakajeka kwatiri ("uyisa jira reusiku"), asi zvakanyanya kuoma kurigadzira sebasa rehunyanzvi. GAN ingori kuedza kutibvisa padambudziko iri.

Pane network mbiri dziri kushanda pano: imwe ijenareta (Generative), imwe inosarura (Adversarial). Imwe network inodzidza kuita basa rinobatsira (kuronga mifananidzo, kuziva manzwi, kudhirowa katuni). Uye imwe network inodzidza kudzidzisa iyo network: ine mienzaniso chaiyo, uye inodzidza kuwana yaimbove isingazivikanwe yakaoma fomula yekuenzanisa kuburitsa kweiyo generative chikamu chemumbure nezvinhu chaizvo-nyika (yekudzidzira muenzaniso) zvichibva pane zvakakosha zvakadzika maficha: nhamba yemeso, kuswedera kune chimiro cheMiyazaki, mataurirwo akarurama eChirungu.

Muenzaniso wekubuda kwetiweki yekugadzira anime mavara. Chinhu

Asi chokwadi, zvakaoma kuvaka chivakwa ipapo. Hazvina kukwana kungokanda neuroni, dzinoda kudzidziswa. Uye zvinotora mavhiki kudzidzisa. Vandinoshanda navo paSamsung AI Center vari kushanda pane iyo GAN musoro, ndeimwe yemibvunzo yavo yakakosha yekutsvagisa. Somuenzaniso, iyi budiriro: kushandisa generative network kugadzira mafoto echokwadi evanhu vane maitiro akasiyana - semuenzaniso, kugadzira kamuri inokwana, kana kugadzira kumeso, izvo zvinogona kuderedza huwandu hweruzivo rwunoda kuchengetwa kana kufambiswa kuti ive nechokwadi chepamusoro-soro yekukurukurirana kwevhidhiyo, kutepfenyura, kana kuchengetedzwa kwekuvanzika.

Chinhu

1.3. Inotsanangurwa AI

Mune mamwe mabasa asingawanzo, kufambira mberi mune zvakadzika zvivakwa zvakangoerekana zvaunza yakadzika neural network padyo nekugona kwevanhu. Ikozvino hondo ndeyekuwedzera huwandu hwemabasa akadaro. Semuyenzaniso, robhoti vacuum cleaner inogona kusiyanisa zviri nyore katsi nembwa pakusangana. Asi mumamiriro ezvinhu ehupenyu chaihwo, hazvizokwanisi kuwana katsi yakarara pakati pemicheka kana fenicha (sezvo, chaizvoizvo, isu tiri, kazhinji ...).

Ndechipi chikonzero chekubudirira kweakadzika neural network? Ivo vanogadzira chinomiririra dambudziko zvichibva kwete paruzivo "runoonekwa neziso rakashama" (pixels emufananidzo, kusvetuka muzwi rezwi, nezvimwewo), asi pazvinhu zvakawanikwa mushure mekugadzirisa ruzivo urwu nemazana akati wandei eiyo neural network. Nehurombo, hukama uhwu hunogonawo kuve husina zvahunoreva, kupokana, kana kutakura zvisaririra zvekusakwana mune yekutanga data set. Semuenzaniso, pane mutambo mudiki wekombuta nezve zvinogona kuitika kana AI ikashandiswa zvisina kufunga mukutsvaga Kupona Kwekunaka Kwakanakisisa.

Mufananidzo wekutaga mifananidzo wakanyorwa kuti munhu ari kubika semukadzi, kunyange zvazvo ari pamufananidzo wacho ari murume (Chinhu) it akacherechedza paVirginia Institute.

Kuti tiongorore hukama hwakaoma uye hwakadzama hwatinowanzotadza kuzvigadzira isu pachedu, tinoda Inotsanangurwa nzira dzeAI. Ivo vanoronga maficha eyakadzika neural network kuitira kuti mushure mekudzidziswa tikwanise kuongorora iyo yemukati inomiririra yakadzidzwa netiweki, uye kwete kungovimba nesarudzo yayo.

1.4. Edge Analytics / AI

Chero chipi chine izwi rekuti Edge chinoreva zvinotevera: kuendesa mamwe maalgorithms kubva pagore / server kuenda kune yekupedzisira mudziyo / gedhi level. Iyo algorithm yakadaro ichashanda nekukurumidza uye haizodi kubatanidza kune yepakati server kushanda. Kana iwe uchiziva nezve kubviswa kwe "mutengi mutete", saka pano isu tiri kuwedzera mutengi uyu zvishoma.
Izvi zvinogona kuve zvakakosha kune Internet yezvinhu. Semuenzaniso, kana muchina wapisa uye uchida kutonhora, zvine musoro kuratidza izvi nekukasira, padanho rekudyara, pasina kumirira kuti data risvike kune gore uye kubva ipapo kuenda kune mutariri wekuchinja. Kana mumwe muenzaniso: mota dzisingatyairi dzinogona kuona mamiriro etraffic ega, pasina kubata sevha yepakati.

Chinhu

Kana mumwe muenzaniso wekuti nei izvi zvakakosha kubva pakuona kwekuchengetedza: paunonyora zvinyorwa pafoni yako, inoyeuka mazwi akajairwa kwauri, kuitira kuti kiibhodhi yefoni yako igone kukupa zano kwauri gare gare - izvi zvinodaidzwa kuti fungidziro yemavara. Kutumira zvese zvaunonyora pane kiibhodhi yako kune imwe nzvimbo kune data data kungave kukanganisa kuvanzika kwako uye kungove kusina kuchengetedzeka. Naizvozvo, kudzidziswa kwekhibhodi kunoitika chete mukati memudziyo wako pachako.

1.5. AI Platform seSevhisi (AI PaaS)

PaaS - Platform seSevhisi - imhando yebhizinesi kwatinowana mukana kune yakasanganiswa chikuva, kusanganisira yayo gore rekuchengetedza data uye akagadzirira-akagadzirwa maitiro. Nekudaro, isu tinokwanisa kuzvisunungura kubva kumabasa ezvivakwa uye kutarisisa zvizere kugadzira chimwe chinhu chinobatsira. Mienzaniso yePaaS mapuratifomu eAI mabasa: IBM Cloud, Microsoft Azure, Amazon Machine Kudzidza, Google AI Platform.

1.6. Adaptive Machine Learning (Adaptive ML)

Zvakadini kana tikabvumira hungwaru hwekugadzira kuti hugadzirise… Unobvunza - ndinoreva, sei?.. Haisi kutojairana nebasa racho here? Dambudziko nderekuti: isu tinogadzira zvine hunyanzvi basa rimwe nerimwe rakadaro tisati tagadzira algorithm yehungwaru yekuigadzirisa. Iwe uchaudzwa - zvinozoitika kuti ketani iyi inogona kurerutswa.

Kudzidza muchina kwakajairika kunoshanda pane yakavhurika-loop musimboti: iwe unogadzirira data, unouya neneural network (kana chero), idzidzise, wobva watarisa mashoma metrics, uye kana iwe uchida zvese, unogona kutumira iyo neural network kune smartphones kugadzirisa matambudziko emushandisi. Asi mumashandisirwo ane data rakawanda uye chimiro chayo zvishoma nezvishoma chinoshanduka, dzimwe nzira dzinodiwa. Masisitimu akadaro anogadzirisa uye anozvidzidzisa akarongwa kuita akavharwa, anozvidzidzira maseketi (akavharwa-loop), uye anofanirwa kushanda zvakanaka.

Zvishandiso zvinogona kuve Stream Analytics, iyo inoshandiswa kuita sarudzo nemabhizinesi mazhinji, kana adaptive kugadzirwa manejimendi. Pachiyero chezvishandiso zvemazuva ano uye uchifunga nezve njodzi dzinonzwisiswa zviri nani kuvanhu, nzira dzinogadzira mhinduro kudambudziko iri, nzira dzese idzi dzinounganidzwa pasi pezita rekuti Adaptive AI.

Chinhu

Kutarisa pamufananidzo uyu, zvakaoma kuzunungusa manzwiro ekuti futurologists hapana chavanoda kunze kwekudzidzisa robhoti kufema ...

Postclassical Compute uye Comms

2.1. Nharembozha yechizvarwa chechishanu (5G)

Iyi inyaya inonakidza zvekuti isu tinobva takuendesa kune yedu chinyorwa. Zvakanaka, heino pfupiso inokurumidza. 5G ichaita kuti iyo Internet ikurumidze zvikuru nekuwedzera iyo data frequency frequency. Masaisai mapfupi ane nguva yakaoma kupfuura zvipingamupinyi, saka iyo network dhizaini ichave yakasiyana zvachose: 500 nguva yakawanda base station inodiwa.

Pamwe chete nekumhanya, isu tichawana zviitiko zvitsva: chaiyo-nguva mitambo ine augmented real, kuita mabasa akaomarara (sekuvhiya) kuburikidza ne telepresence, kudzivirira tsaona uye mamiriro akaoma munzira kuburikidza nekutaurirana pakati pemotokari. More prosaic: pakupedzisira, nharembozha inomira kudonha panguva yezviitiko zvakawanda, senge mutambo munhandare.

Mufananidzo kunobva: Reuters, Niantic

2.2. Chinotevera-chizvarwa Memory

Pano tiri kutaura nezve chizvarwa chechishanu che RAM - DDR5. Samsung yakazivisa kuti pakupera kwa2019, DDR5-yakavakirwa zvigadzirwa zvichaonekwa. Iyo ndangariro nyowani inotarisirwa kunge ichimhanya zvakapetwa kaviri uye nekaviri secapacious tichichengeta fomu rechinhu, zvichireva kuti tichakwanisa kuwana matanda ekurangarira ane huwandu hunosvika 32 GB kumakomputa edu. Mune ramangwana, izvi zvichanyanya kukosha kune smartphones (iyo nyowani ndangariro ichave mune yakaderera-simba vhezheni) uye yelaptops (uko huwandu hweDIMM slots hushoma). Kudzidza kwemichina kunodawo huwandu hukuru hwe RAM.

2.3. Yakaderera-Earth-Orbit Satellite Systems

Pfungwa yekutsiva inorema, inodhura, ine simba setiraiti ine pungwe diki uye yakachipa iri kure nekutsva uye yakaonekwa kumashure kuma90s. Nezvekuti "Elon Musk achakurumidza kugovera Internet kumunhu wese kuburikidza ne satellite" Mazuvano, simbe chete hadzisati dzanzwa nezvazvo. Kambani ine mukurumbira pano iIridium, iyo yakashaya mari mukupera kwe90s asi yakaponeswa neUS Department of Defense (kwete kuti ivhiringike neRidium, yeRussia smart home system). Elon Musk's project (Starlink) iri kure neyoga - Richard Branson (OneWeb - 1440 yakarongwa satellites), Boeing (3000 satellites), Samsung (4600 satellites), uye vamwe vari kutora chikamu mujaho resatellite.

Mamiriro akaita sei munzvimbo ino, hupfumi hwakamira sei ikoko - verenga mukati review. Uye isu takamirira bvunzo dzekutanga dzeaya masisitimu nevashandisi vekutanga, izvo zvinofanirwa kuitika gore rinouya.

2.4. Nanoscale 3D Printing

3D kudhinda, kunyangwe isina kupinda muhupenyu hwemunhu wese (muchimiro cheakavimbiswa ega epurasitiki fekitori yemumba), zvakadaro zvakapfuura kare niche yetekinoroji ye geeks. Tichifunga nezvekuti chero mwana wechikoro anoziva nezvekuvapo kweinenge 3D-sculptural pens, uye vazhinji vanorota kutenga bhokisi nevamhanyi uye extruder ye ... "saizvozvo" (kana vatotenga imwe).

Stereolithography (laser 3D printers) inobvumira kudhinda nemafotoni ega: mapolymer matsva ari kudzidzwa anoda mafotoni maviri chete kuti aomese. Izvi zvinobvumira mafirita matsva, zvisungo, zvitubu, macapillaries, lenzi uye ... sarudzo dzako mumashoko kuti dzigadzirwe mumamiriro asiri emurabhoritari! Uye haisi kure ne photopolymerization - iyi teknolojia chete inobvumira "kudhinda" processors uye computing circuits. Mukuwedzera, kwemakore akawanda ikozvino tekinoroji yekudhinda graphene 500nm XNUMXD zvimiro, asi pasina kukura kukuru.

Chinhu

3. Sensing uye Mobility

3.1. Autonomous Driving Level 4 & 5

Kuti urege kuvhiringika muizwi, zvakakodzera kunzwisisa kuti ndeapi mazinga ekuzvimiririra anosiyaniswa (akatorwa kubva kune yakadzama zvinyorwa, kwatinoreva vese vanofarira):

Level 1: Cruise control: inobatsira mutyairi mumamiriro mashoma kwazvo (semuenzaniso kuchengeta mota iri kumhanya yakatarwa mushure mekunge mutyairi abvisa tsoka yake padanda)
Chikamu chechipiri: Rubatsiro rwekutungamirira uye bhureki shoma. Mutyairi anofanira kunge akagadzirira kutora chigaro nekukurumidza. Maoko ake ari pavhiri, maziso ake ari munzira. Izvi ndizvo izvo Tesla naGeneral Motors vatove nazvo.
Level 3: Mutyairi haachafaniri kugara akatarisa mugwagwa. Asi unofanira kugara wakangwarira uye kugadzirira kudzora. Ichi chinhu chisati chawanikwa mumotokari dzinotengeswa. Vese varipo vari palevel 1-2.
Nhamba 4: Kuzvifambisa kwechokwadi, asi nemiganhu: kungotyaira munzvimbo inozivikanwa yakanyatsorongwa uye inowanzozivikanwa kune system, uye pasi pemamwe mamiriro: semuenzaniso, hapana chando. Waymo naGeneral Motors vane prototypes yeizvi, uye vanoronga kuamisa mumaguta akati wandei uye kuvayedza mumamiriro ezvinhu epasirese. Yandex ine nzvimbo dzekuyedza dzematekisi asingatyairi muSkolkovo neInnopolis: kukwira kunoitika pasi pekutariswa kweinjiniya agere pachigaro chevatakuri; kambani inoronga kuwedzera zvikepe zvayo kusvika kumotokari zana dzisingatyairi pakupera kwegore.
Level 5: Yakazara otomatiki kutyaira, kutsiva mutyairi wemunhu zvachose. Maitiro akadaro haapo uye haafanire kuoneka mumakore anotevera.

Ndezvechokwadi sei kuona zvose izvi mune ramangwana rinofanoona? Pano ndinoda kudzosera muverengi kuchinyorwa "Nei Tesla Achivimbisa Kutangisa Robotaxis na2020 Hazvigoneke". Izvi zvinokonzerwa nekushaikwa kwe5G yekubatanidza: ikozvino 4G kumhanya hakuna kukwana. Zvimwe nekuda kwemutengo wakakwira kwazvo wemashini anozvimiririra: ivo havasati vaita purofiti, iyo bhizinesi modhi haina kujeka. Muchidimbu, "zvese zvakaoma" pano, uye hazvina kuitika kuti Gartner anonyora kuti fungidziro yekushandiswa kwehuwandu hweNhamba 4 ne5 haina kutanga kupfuura mumakore gumi.

3.2. 3D Sensing Makamera

Makore masere apfuura, Microsoft's Kinect yemitambo controller yakaita kupenga nekupa inosvikika uye isingadhure mhinduro ye3D chiratidzo. Kubva ipapo, Kinect-yakavakirwa padzidzo yemuviri nemitambo yekutamba yakave nekusimuka nekudonha kwenguva pfupi, asi 3D makamera akashandiswa mumarobhoti eindasitiri, mota dzisina munhu, uye nharembozha kuti dzizivikanwe kumeso. Iyo tekinoroji yave yakachipa, yakawedzera compact, uye inowanikwa nyore.

Iyo Samsung S10 foni ine Nguva-ye-Ndege kamera inoyera chinhambwe kuenda kuchinhu kuita kuti kutarisa kuve nyore. Chinhu

Kana iwe uchifarira nyaya iyi, isu tinokuendesa kune yakanaka kwazvo yakadzama wongororo yeakadzama kamera: chikamu 1, chikamu 2.

3.3. Light Cargo Delivery Drones

Amazon yakaita ruzha gore rino payakafumura dhiraivha nyowani inokwanisa kutakura mitoro midiki inorema kusvika 2 kg. Kuguta, nekumanikidzana kwaro kwemotokari, izvi zvinoita semhinduro yakanaka. Ngationei kuti ma drones aya achaita sei munguva pfupi iri kutevera. Zvichida zvakakosha kusanganisira kungwarira kusava nechokwadi pano: kune matambudziko mazhinji, kubva pamukana wekuba nyore kwe drone kusvika kune zvirambidzo zvemutemo paUAVs. Amazon Prime Air yanga iripo kwemakore matanhatu, asi ichiri muchikamu chekuyedza.

Drone nyowani yeAmazon, yakavhurwa chitubu ichi, ine zvishoma zveStar Wars kunzwa kwairi. Chinhu

Kunze kweAmazon, kune vamwe vatambi mumusika uyu (pane yakadzama ongorora), asi kwete chigadzirwa chimwe chete chakapedzwa: zvese zviri padanho rekuyedza uye rekushambadzira. Zvakakosha kucherechedza zvakasiyana izvo zvinonyanya kunakidza zvekurapa nyanzvi zvirongwa muAfrica: kupa ropa kuendeswa kuGhana (14 kuendeswa, Zipline) neRwanda (Matternet).

3.4. Flying Autonomous Vehicles

Zvakaoma kutaura chero chinhu chaicho pano. Sekureva kwaGartner, izvi hazvizoonekwa zvisati zvaitika mumakore gumi. Kazhinji, heino matambudziko ose akafanana mumotokari dzisina mutyairi, ivo chete vanowana hutsva hutsva - vertical. Porsche, Boeing, Uber vanozivisa vavariro yavo yekuvaka tekisi inobhururuka.

3.5. Augmented Reality Cloud (AR Cloud)

Iyo yekusingaperi kopi yedhijitari yenyika chaiyo, ichibvumira kuumba hutsva hwechokwadi, hunozivikanwa kune vese vashandisi. Mune mamwe mazwi ehunyanzvi, ndeye kugadzira yakavhurika gore chikuva, umo vagadziri vanogona kubatanidza yavo AR maapplication. Iyo monetization modhi yakajeka, imhando yeanalogue yeSteam. Pfungwa iyi yakadzika midzi zvekuti ikozvino vamwe vanhu vanotenda kuti AR isina gore haina basa.

Kuti zvingaratidzika sei mune ramangwana zvinoratidzwa muvhidhiyo pfupi. Inotaridzika sechimwe chikamu che "Black Mirror":

Unogonawo kuverenga mukati ongororo chinyorwa.

4. Augmented Human

4.1. Emotion AI

Nzira yekuyera sei, kutevedzera uye kupindura kune manzwiro evanhu? Vamwe vevatengi pano makambani anogadzira vabatsiri vezwi seAmazon Alexa. Vachakwanisa kunyatsozvinyudza mudzimba kana vakadzidza kuziva mafungiro: kunzwisisa chikonzero chekusagutsikana kwemushandisi, edza kugadzirisa mamiriro acho ezvinhu. Kazhinji, kune ruzivo rwakawanda muchirevo kupfuura mumharidzo pachayo. Uye chirevo chechiso chitarisiko chechiso, matauriro, uye kusataura.

Zvimwe zvinoshandiswa zvinoshanda zvinosanganisira: kuongorora manzwiro panguva yekubvunzurudzwa kwebasa (uchishandisa vhidhiyo kubvunzurudzwa), kuongorora maitiro kune zvekutengesa kana zvimwe zvevhidhiyo zvemukati (kunyemwerera, kuseka), uye kubatsira nekudzidza (semuenzaniso, yekudzidzira yakazvimirira muhunyanzvi hwekutaura pachena).

Zvakaoma kutaura pamusoro penyaya iyi zvirinani pane munyori wefirimu remaminitsi matanhatu Kuba Manzwiro Edu. Vhidhiyo yakangwara uye yakashongedzwa inotaridza kuti manzwiro edu anogona kuyerwa sei nechinangwa chekushambadzira, uye kubva pakuita kwechiso chako, unogona kuziva: unofarira pizza, imbwa, Kanye West, uye kunyangwe nhanho yako yemuhoro uye inofungidzirwa IQ. Nekudzvanya pane chinongedzo chiri pamusoro, iwe unotora chikamu muvhidhiyo inopindirana uchishandisa laptop yako yakavakirwa-mukati kamera. Iyi firimu yakatoratidzwa pamitambo yakawanda yemafirimu.

Chinhu

Kune kunyange chidzidzo chinonakidza chakadaro: maitiro ekuziva sarcasm mune zvinyorwa. Vakatora ma tweets ane hashtag #sarcasm uye vakaita sampuli yekudzidzira ye25 tweets nekunyomba uye 000 nguva dzose tweets pamusoro pezvinhu zvose pasi pezuva. Vakashandisa raibhurari yeTensorFlow, vakadzidzisa sisitimu, uye heino mhedzisiro:

Chinhu

Saka ikozvino, kana iwe usina chokwadi nezve waunoshanda naye kana shamwari - kunyangwe akakuudza chimwe chinhu zvakakomba kana kunyomba - unogona kushandisa zvagara kudzidziswa neneural network!

4.2. Augmented Intelligence

Automation yebasa rehungwaru uchishandisa nzira dzekudzidza muchina. Hapana chitsva, zvingaite? Asi iwo mazwi pachawo akakosha pano, kunyanya nekuti anopindirana muchidimbu neArtificial Intelligence. Izvi zvinotinongedzera kune gakava pamusoro pe "yakasimba" uye "isina simba" AI.
Yakasimba AI ihwo hungwaru hwekunyepedzera kubva kumafirimu engano dzesainzi akaenzana zvachose nepfungwa dzemunhu uye anozviziva ari munhu. Izvi hazvisati zvavepo uye hazvizivikanwe kana zvichavepo zvachose.

Weak AI haisi hunhu hwakazvimirira, asi mubatsiri wemunhu. Haizviti ine kufunga kwakafanana nevanhu, asi inongoziva nzira yekugadzirisa matambudziko eruzivo, semuenzaniso, kuona zvinoratidzwa mumufananidzo kana kushandura zvinyorwa.

Chinhu

Nenzira iyi, Augmented Intelligence yakachena "AI isina simba," uye mazwi anoita seakakodzera, sezvo asingaunzi nyonganiso kana muedzo wekuona pano iyo "yakasimba AI" iyo munhu wese anorota (kana kutya, kana tikarangarira hurukuro dzakawanda nezve "kusimuka kwemichina"). Tichishandisa chirevo cheAugmented Intelligence, tinobva tangova magamba erimwe firimu: kubva mune zvesainzi ngano (seAsimov's "Ini, Robot") tinozviwana tiri mucyberpunk ("kuwedzeredza" mumhando iyi marudzi ese ezvinomera zvinowedzera kugona kwevanhu).

sei akadaro Erik Brynjolfsson naAndrew McAfee: "Izvi ndizvo zvichaitika mumakore gumi anotevera. Hazvirevi kuti AI ichatsiva vatungamiri, asi kuti vatungamiri vanoshandisa AI vachatsiva avo vasati vasvika."

mienzaniso:

• Mushonga: Stanford University yakagadzira algorithm, uyo anorarama nebasa rekuziva pathologies pachipfuva X-rays paavhareji sekubudirira sevanachiremba vazhinji.
• Dzidzo: rubatsiro kuvadzidzi nevadzidzisi, kuongororwa kwemhinduro dzevadzidzi kune zvigadzirwa, kuvandudza nzira dzekudzidza dzega.
• Business analytics: pre-processing yedata, maererano nehuwandu, inotora 80% yenguva yemuongorori, uye chete 20% ndiyo yekuedza pachayo.

4.3. Biochips

Iri ndiro dingindira rinodiwa reese cyberpunk mafirimu nemabhuku. Kazhinji, kucheka mhuka dzinovaraidza haisi tsika itsva. Asi ikozvino machipisi aya atanga kuiswa muvanhu zvakare.

Muchiitiko ichi, iyo hype inogona kunge yakabatana neyepamusoro-mbiri kesi muAmerican kambani Three Square Market. Ikoko, mushandirwi akatanga kugovera machipisi pasi peganda kuti apiwe muripo. Iyo chip inokubvumira kuti uvhure magonhi, pinda mumakomputa, tenga chikafu chekudya kubva kumuchina wekutengesa - ndiko kuti, kadhi yevashandi vepasirese. Panguva imwecheteyo, chip yakadaro inoshanda chaizvo sekadhi rekuzivikanwa, haina GPS module, saka hazvigoneke kuteedzera chero munhu nayo. Uye kana munhu achida kubvisa chip kubva muruoko rwake, zvinotora maminitsi mashanu nerubatsiro rwechiremba.

Machipisi anowanzo kuisirwa pakati pechigunwe chechigunwe nechigunwe. Chinhu

Verenga zvimwe chinyorwa nezve mamiriro ezvinhu ne microchipping munyika.

4.4 Immersive Workpace

“Kunyura” nderimwe izwi idzva risingagoni kunzvengwa. Riri kwese kwese. Immersive theatre, kuratidzwa, cinema. Zvinorevei? Kunyura ndiko kusikwa kwemaitiro anonyudza, kana muganho pakati pemunyori nemuoni, iyo chaiyo uye nyika chaiyo yakarasika. Munzvimbo yebasa, zvinofanirwa kufungidzirwa kuti izvi zvinoreva kudzima muganho pakati pemuiti neanotanga uye nekukurudzira vashandi kuti vatore chinzvimbo chekushanda nekugadzirisa nharaunda yakavapoteredza.

Sezvo isu tave neAgile, kuchinjika, uye kubatana kwepedyo kwese kwese, nzvimbo dzebasa dzinofanirwa kugadzirika zviri nyore sezvinobvira uye dzinofanira kukurudzira kushanda kwechikwata. Upfumi hunoraira mamiriro ayo: kune vashandi venguva pfupi, mari yekurenda nzvimbo yehofisi iri kukura, uye mumamiriro emakwikwi emusika wevashandi muIT, makambani ari kuedza kuwedzera kugutsikana kwevashandi kubva kubasa nekugadzira nzvimbo dzekuvaraidza uye zvimwe zvinobatsira. Uye izvi zvese zvinoratidzwa mukugadzirwa kwenzvimbo dzebasa.

Of chirevo knoll

4.5. Munhu

Wese munhu anoziva chinonzi personalization mukushambadzira. Ndiyo apo iwe uri kukurukura nemumwe waunoshanda naye nhasi kuti mhepo iri mukamuri yakaoma uye iwe unofanira kutenga humidifier yehofisi, uye zuva rinotevera unoona chiziviso pasocial network yako - "tenga humidifier" (nyaya chaiyo yakaitika kwandiri).

Chinhu

Kusarudzika, maererano naGartner, imhinduro kune kushushikana kuri kukura kwevashandisi nezve mashandisiro e data ravo rega nekuda kwezvinangwa zvekushambadzira. Chinangwa ndechekugadzira nzira yaticharatidzwa kushambadzira inoenderana nemamiriro atiri, uye kwete kwatiri isu pachedu. Semuenzaniso, nzvimbo yedu, rudzi rwechigadzirwa, nguva yezuva, mamiriro ekunze - ichi chinhu chisingaputse data yedu yega, uye hatinzwi kunzwa kusingafadzi kwe "kutevera".

Verenga pamusoro pemusiyano wepfungwa mbiri idzi chinyorwa Andrew Frank mune blog paGartner webhusaiti. Pane mutsauko wakadzikama uye mazwi akadaro akafanana zvekuti, usingazivi mutsauko, unoisa pangozi yekukakavadzana nemutauriri wako kwenguva yakareba, usingafungi kuti, kazhinji, zvose zviri zviviri vakarurama (uye iyi inyaya chaiyo yakaitika kune munyori).

4.6. Biotech - Yakagadzirwa kana Artificial Tissue

Iyi ndiyo, yekutanga yezvose, pfungwa yekukura nyama yakagadzirwa. Zvikwata zvakati wandei kutenderera pasirese zviri kushanda panguva imwe chete kugadzira marabhoritari "Nyama 2.0" - inotarisirwa kuve yakachipa kupfuura zvakajairwa, uye maresitorendi echikafu nekukurumidza uyezve masupamaketi achachinja kwairi. Vanoisa mari mune iyi tekinoroji vanosanganisira Bill Gates, Sergey Brin, Richard Branson nevamwe.

Chinhu

Zvikonzero nei munhu wese achifarira nyama yekugadzira:

1. Global warming: methane emissions kubva kumapurazi. Ichi chikamu che18% chegasi rinoshandura mamiriro ekunze pasi rose.
2. Kuwedzera kwevanhu. Kudiwa kwenyama kuri kukura, uye hazvibviri kudyisa munhu wose ane nyama yechisikigo - inongodhura.
3. Kushaya nzvimbo. 70% yesango reAmazon rakatotemerwa mafuro.
4. Pfungwa dzetsika. Kune vakakoshera izvi. Sangano rinorwira kodzero dzemhuka rePETA ratopa mubairo wemiriyoni imwe chete yemadhora kunyanzvi yesainzi ichaunza nyama yehuku kumusika.

Kutsiva nyama chaiyo nenyama yesoya mhinduro ine chikamu, nekuti vanhu vanonzwa mutsauko wekuravira uye magadzirirwo zvakanaka, uye havagone kusiya steak vachifarira soy. Saka chaiyo, nyama yakasimwa inodiwa. Iye zvino, zvinosuruvarisa, nyama yekugadzira inodhura zvakanyanya: kubva pamadhora gumi nemaviri pakirogiramu. Izvi zvinokonzerwa neiyo yakaoma tekinoroji maitiro ekukura nyama yakadaro. Verenga pamusoro pezvose izvi chinyorwa.

Kana tikataura nezve mamwe mamiriro ekukura kwematishu - atove mukurapa - ipapo musoro wenhengo dzekugadzira unonakidza: semuenzaniso, "chigamba" chemhasuru yemoyo, rakadhindwa ine yakakosha 3D printer. Ivo vanozivikanwa nhoroondo senge moyo wembeva wakakura, asi zvakazara, zvese hazvisati zvaenda kupfuura miedzo yekiriniki. Saka isu hatigone kuona Frankenstein mumakore anotevera.

Pano, Gartner anongwarira muongororo yake, sezviri pachena achirangarira fungidziro yake yakakundikana ye2015 kuti muna 2019, 10% yehuwandu hwenyika dzakabudirira vachange vaine 3D-yakadhindwa yekurapa yekumisikidza. Ndosaka achidoma nguva yekusvika padanho rekubereka isingasviki makore gumi.

5. Digital Ecosystems

5.1. Decentralized Web

Pfungwa iyi inodyidzana zvakanyanya nezita remuvambi wewebhu, Turing Award anohwina Sir Tim Burners-Lee. Aigara achifunga nezvehutsika hwesainzi yekombuta uye kubatana kweInternet: paakaisa hwaro hwe hypertext, aive nechokwadi chekuti network inofanira kushanda sewebhu, kwete sehumambo. Izvi zvaive zvechokwadi mumazuva ekutanga enetwork. Nekudaro, sezvo iyo Internet yakakura, chimiro chayo chakava nechepakati nekuda kwezvikonzero zvakati. Zvakazoitika kuti kupinda kune network kune nyika yese kunogona kuvharika zviri nyore uchishandisa vashoma vanopa. Uye data remushandisi rakava sosi yesimba uye mari kumakambani eInternet.

"Indaneti yakatove yakasarudzika," anodaro Burners-Lee. "Dambudziko nderekuti injini yekutsvaga, imwe hombe yesocial network, imwe microblogging platform inotonga. Isu hatina dambudziko rehunyanzvi, asi isu tine dambudziko remagariro."

Mune rake tsamba yakazaruka Pamakore makumi matatu ekuyeukwa kweWorld Wide Web, mugadziri weWebhu akatsanangura matambudziko matatu makuru neInternet:

1. Kukuvadza kwakanangana, sekubira kunotsigirwa nehurumende, kuparwa kwemhosva, uye kushungurudza pamhepo
2. Iwo chaiwo magadzirirwo ehurongwa, izvo, mukukanganisa kwemushandisi, zvinogadzira nzvimbo yemaitiro akadai se: mari yekukurudzira yeclickbait uye kupararira kwehutachiona hwemashoko enhema.
3. Zvisingatarisirwe mhedzisiro yedhizaini dhizaini inotungamira mukukonana uye yakashata mhando yekukurukurirana online

Uye Tim Berners-Lee atove nemhinduro yekuti ndeapi misimboti iyo "Healthy Internet" inogona kuve yakavakirwa, isina dambudziko nhamba 2: "Kune vashandisi vazhinji, iyo chete modhi yekudyidzana newebhu imari yekushambadzira. Kunyangwe kana vanhu vachitya nezve zviri kuitika kune yavo data, ivo vakagadzirira kuita chibvumirano nemuchina wekushambadzira mukana wekuwana zvemukati zvemahara uye kubhadhara kuri nyore kwepasirese uye kubhadhara kuri nyore kwesevhisi. mapato maviri." Pakati pesarudzo dzekuti izvi zvinogona kurongwa sei: vaimbi vaigona kutengesa marekodhi avo vasina vamiriri vakaita se iTunes, uye masaiti enhau anogona kushandisa micropayment system yekuverenga chinyorwa chimwe chete, pane kuwana mari pakushambadzira.

Semuenzaniso wekuyedza weiyi Internet nyowani, Tim Berners-Lee akatanga chirongwa cheSOLID, pfungwa iri yekuti iwe unochengeta data rako mu "pod" - imba yekuchengetera data - uye inogona kupa iyo data kune yechitatu-bato zvikumbiro. Asi musimboti, iwe ndiwe tenzi we data rako. Izvi zvese zvine hukama zvakanyanya neiyo pfungwa yepeer-to-peer network, zvichireva kuti komputa yako haingokumbire masevhisi, asi zvakare inovapa, kuti urege kuvimba neserver imwe seyo yega chiteshi.

Chinhu

5.2. Decentralized Autonomous Organizations

Iri isangano rinotungamirirwa nemitemo yakanyorwa nenzira yepurogiramu yekombiyuta. Basa rayo rezvemari rinobva pane blockchain. Chinangwa chekugadzira masangano akadai ndechekubvisa hurumende kubva pachinzvimbo chemurevereri uye kugadzira nharaunda yakavimbika inovimbwa kune vanopikisa, iyo isiri yemunhu chero upi zvake, asi ndeyemunhu wese pamwechete. Ndiko kuti, muchirevo, izvi zvinofanirwa, kana zano racho rikadzika midzi, bvisa vanyori uye mamwe masangano echinyakare ekusimbisa.

Muenzaniso unonyanya kuzivikanwa wesangano rakadaro waive wekutsvaga DAO, iyo yakasimudza $ 2016 mamiriyoni muna 150, iyo $ 50 mamiriyoni yakabva yabiwa pakarepo kuburikidza nemutemo wepamutemo mumitemo. Dambudziko rakaoma rakamuka pakarepo: kana kukungurutsa uye kudzorera mari, kana kubvuma kuti kubviswa kwemari kwaiva pamutemo, nokuti hazvina kutyora mitemo yepuratifomu. Somugumisiro, kuitira kuti vadzorere mari kune vanoisa mari, vagadziri vaifanira kuparadza The DAO nekunyorazve blockchain uye kutyora musimboti wayo mukuru - kusachinja.

Comic nezve Ethereum (kuruboshwe) uye Iyo DAO (kurudyi). Chinhu

Iyi nyaya yese yakaparadza mukurumbira weiyo DAO pfungwa pachayo. Icho chirongwa chakanga chakavakirwa paEthereum cryptocurrency, uye Ethereum 2.0 inotarisirwa gore rinouya - zvichida vanyori (kusanganisira yakakurumbira Vitalik Buterin) vachafunga nezvezvikanganiso uye kuratidza chimwe chinhu chitsva. Ndosaka Gartner akaisa DAO pamutsetse unokwira.

5.3.Synthetics Data

Kudzidzisa neural network kunoda huwandu hukuru hwe data. Kunyora data nemaoko ibasa rakakura kwazvo rinogona kuitwa nemunhu chete. Saka, zvigadzirwa zve data zvekugadzira zvinogona kugadzirwa. Semuenzaniso, kuunganidzwa kwakafanana kwezviso zvevanhu pawebhusaiti https://generated.photosIwo akagadzirwa achishandisa GAN algorithms, ayo anga atotaurwa pamusoro apa.

Zviso izvi hazvisi zvevanhu. Chinhu

Kubatsira kukuru kwedata rakadaro nderekuti hapana matambudziko emutemo mukuishandisa: hapana munhu anopa mvumo yekugadziriswa kwedata rako.

5.4.Digital Ops

Iyo suffix "Ops" yave yakajairika kubvira DevOps yatora midzi mukutaura kwedu. Zvino pamusoro pekuti DigitalOps chii - ingori ruzhinji rweDevOps, DesignOps, MarketingOps… Watobhowekana here? Muchidimbu, iko kuendeswa kweiyo DevOps nzira kubva kune software sphere kune mamwe ese maficha ebhizinesi - kushambadzira, dhizaini, nezvimwe.

Chinhu

Pfungwa yeDevOps yaive yekubvisa zvipingamupinyi pakati peBudiriro neKushanda (maitiro ebhizinesi) nekugadzira zvikwata zvakajairika, kusanganisira vanogadzira mapurogiramu, vanoedza, nyanzvi dzekuchengetedza, uye vatariri; kushandisa mamwe maitiro: kuenderera mberi kubatanidzwa, zvivakwa sekodhi, kupfupisa nekusimbisa mhinduro loops. Chinangwa chaive chekumhanyisa nguva yekuenda kumusika. Kana iwe waifunga kuti izvi zvakafanana neAgile, waifunga nemazvo. Iye zvino fambisa pfungwa iyi kubva kune software yekuvandudza sphere kuenda kubudiriro mune zvakajairika - uye iwe unonzwisisa kuti DigitalOps chii.

5.5. Ruzivo Magirafu

Iyo software nzira yekuenzanisira ndima yeruzivo, kusanganisira kushandisa muchina kudzidza algorithms. Girafu yeruzivo inovakwa pamusoro peiyo iripo dhatabhesi yekubatanidza pamwechete ruzivo rwese: zvese zvakarongwa (rondedzero yezviitiko kana vanhu) uye isina kurongeka (chinyorwa chechinyorwa).

Muenzaniso wakapfava ndewekadhi raunoona mumhinduro dzekutsvaga paGoogle. Kana ukatsvaga munhu kana sangano, uchaona kadhi kurudyi:


Ndapota cherechedza kuti "Zviitiko Zvinouya" haisi kopi yemashoko kubva kuGoogle Maps, asi kubatanidzwa kwepurogiramu kubva kuYandex.Afisha: unogona kuona izvi nyore kana ukanyora pazviitiko. Ndiko kuti, iko kusanganiswa kweakawanda data masosi.

Kana iwe ukakumbira runyoro - semuenzaniso, "vatungamiriri vane mukurumbira" - iwe ucharatidzwa "carousel":

Bonus kune avo vanoverenga kusvika kumagumo

Uye zvino, kana tazvijekesa pachedu zvinorehwa neimwe yepfungwa, tinogona kutarisa mufananidzo wakafanana, asi muchiRussia:

Inzwa wakasununguka kuigovera pasocial network!

Tatyana Volkova - Munyori wekharikhulamu yeInternet Yezvinhu track paSamsung IT Academy, nyanzvi muzvirongwa zvemakambani ekutarisira paSamsung Research Center.


Source: www.habr.com