Page 14 of 24 FirstFirst ... 411121314151617 ... LastLast
Results 261 to 280 of 463

Thread: Omul si computerul

  1. Back To Top | #261
    Moderator Ivan's Avatar
    Join Date
    Mar 2001
    Location
    Bucuresti
    Ii stricati de tot feng shuiul lui Oldshooterman care e convins ca undeva ascuns intr-o debara e un AI constient si cu liber arbitru.

  2. Back To Top | #262
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    Nu e constient, dar are liberul arbitru. Daca "A" il invata un lucru, "B" il invata alt lucru iar "C" il intreaba ce a invatat, raspunsul nu e preprogramat, precum s-a demonstrat. Tay a spus ce nu le-a convenit, DeepMind a batut campionii la GO printr-o strategie proprie, altul a devenit campion la un joc invatat de la zero, fara ca nici macar sa cunoasca regulile jocului. Conceptul ca un program nu poate fi inteligent pentru ca un om l-a creat e depasit.

    Artificial Intelligence: Can a Machine Think?
    .........................................................

    "A senior Google engineer AI accepted the United States for science and technology reporter an exclusive interview to the competition and the prospect of artificial intelligence published his views. Since the interview without permission, he can not disclose the specific identity."(geksake)

    Q: - How do you see the possibility of the existence of such a Skynet?
    R: - I think that with the advent of artificial general intelligence (AGI) era, this is possible. But the capacity of the current artificial intelligence is too limited and too far away from this possibility further. I estimate that 10 years later the problem will become possible.

    Cam tot peste 10 ani ar fi trebuit sa fie si victoria la GO.
    Last edited by oldshooterman; 28-03-2016 at 10:10.

  3. Back To Top | #263
    Moderator Ivan's Avatar
    Join Date
    Mar 2001
    Location
    Bucuresti
    Quote Originally Posted by oldshooterman View Post
    Nu e constient, dar are liberul arbitru.
    Tot e un pas inainte.

    Acum sa cadem de acord asupra definitiei liberului arbitru.

  4. Back To Top | #264
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    Liberul arbitru = libertatea omului de a alege între posibilitățile care i se oferă. E valabil si pentru AI, demonstrat recent de catre AlpgaGo prin "Mutarea 37", de ex..

    "For Move 37, the probability was one in ten thousand. In other words, AlphaGo knew this was not a move that a professional Go player would make."

    "But, drawing on all its other training with millions of moves generated by games with itself, it came to view Move 37 in a different way. It came to realize that, although no professional would play it, the move would likely prove quite successful. “It discovered this for itself,” Silver says, “through its own process of introspection and analysis."(David Silver, the guy who led the creation of AlphaGo)

    Adica "informatia" se descurca singura in anumite situatii. Chiar daca tie nu-ti place ideea, la multi le place, mie chiar in mod deosebit. E un inceput promitator.

  5. Back To Top | #265
    Moderator Ivan's Avatar
    Join Date
    Mar 2001
    Location
    Bucuresti
    Iar incepem cu science fiction pentru gimnaziu si wishful thinking.

    Stiu ca tie iti place ideea, nu stiu de ce crezi ca mie nu (pentru ca nu am spus vreodata nici ca imi place, nici ca nu-mi place).
    Ce nu pari sa pricepi este ca pentru un eveniment oarecare nu conteaza daca iti place sau dimpotriva, iti repugna. Evenimentul respectiv va avea loc sau nu dupa cu totul alte criterii.

    Nu-l ajuta ca ii faci lobby, dar ce-i drept nici nu-l deranjeaza.

  6. Back To Top | #266
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    LeCun despre cand crede ca se va sti cum functioneaza inteligenta umana: "“We kind of know the ingredients; we just don’t know the recipe,” he says. “It might take a while.”

    Solutia pentru AI e "invatarea nesupravegheata". Deocamdata a trecut de la faza primara cu "deep learning" la o faza intermediara cu "deep learning + reinforcement learning (software is trained by receiving automatic feedback on its performance)". Se experimenteaza chiar si cu forme limitate de "invatare nesupravegheata".

    Dar liberul arbitru e chiar la baza ideii. Poti intreba un bot mai istet ce fel de flori iti recomanda, intre lalele si trandafiri de ex., fara sa stie ce-ti place. Nu-i exclus nici sa-ti recomande altele care-i plac lui. Oricare solutie o alege demonstreaza ca liberul arbitru exista in luarea deciziilor.

  7. Back To Top | #267
    Moderator Ivan's Avatar
    Join Date
    Mar 2001
    Location
    Bucuresti
    Quote Originally Posted by oldshooterman View Post
    Oricare solutie o alege demonstreaza ca liberul arbitru exista in luarea deciziilor.
    Te inseli.
    Daca nu ma crezi citeste un pic despre generarea unui numar aleator folosind un computer.

    Daca avand aceleasi conditii initiale va alege diferit - da, putem vorbi despre liberul arbitru. Adica putem aduce vorba, nu inseamna ca musai exista (din cate stiu nu exista un consens privind liberul arbitru in general, daramite la masini).

  8. Back To Top | #268
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    AI and Free Will: When do choices exist? (David McAllester/computer scientist)

    "My conclusion is that choices exist even when the decision making process is deterministic."

    "Summary: It seems to me that "free will" cannot be subjectively distinguished from having choices. And we do have choices --- like the chess program we must still choose, even if we are determined and predictable."

  9. Back To Top | #269
    Moderator Ivan's Avatar
    Join Date
    Mar 2001
    Location
    Bucuresti
    Asta deja e o problema de semantica.
    Daca sunt "determined and predictable" IMO nu sunt alegeri.

  10. Back To Top | #270
    Paranoia is natural AStateOfMind's Avatar
    Join Date
    Nov 2000
    Location
    Portland, OR
    Liberul arbitru=posibilitatea ca atunci cand esti supus unor reguli si teorii sa-ți bagi plua si sa faci ceva aberant. Nu vreau ai cu liber arbitru.
    "Everyone believes in the atrocities of the enemy and disbelieves in those of his own side, without ever bothering to examine the evidence.”

  11. Back To Top | #271
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    Lucrurile se misca in directia buna: "...it’s probably going to be a few years before we begin to see something like it on the market.":

    IBM Invents 'Resistive' Chip That Can Speed Up AI Training By 30,000x

    "However, even Google’s AlphaGo still needed thousands of chips to achieve its level of intelligence. IBM researchers are now working to power that level of intelligence with a single chip, which means thousands of them put together could lead to even more breakthroughs in AI capabilities in the future."

    Cine cauta bine, pana la urma gaseste.

  12. Back To Top | #272
    Geoffrey Hinton, the godfather of deep learning despre AI, Alphago, The universe and everything else:

    Geoffrey Hinton, the 'godfather' of deep learning, on AlphaGo

    Q: Beyond games, then—what might come next for AI?
    A: It depends who you talk to. My belief is that we’re not going to get human-level abilities until we have systems that have the same number of parameters in them as the brain. So in the brain, you have connections between the neurons called synapses, and they can change. All your knowledge is stored in those synapses. You have about 1,000-trillion synapses—10 to the 15, it’s a very big number. So that’s quite unlike the neural networks we have right now. They’re far, far smaller, the biggest ones we have right now have about a billion synapses. That’s about a million times smaller than the brain.

    Q: Do you dare predict a timeline for that?
    A: More than five years. I refuse to say anything beyond five years because I don’t think we can see much beyond five years. And you look at these past predictions like there’s only a market in the world for five computers [as allegedly said by IBM founder Thomas Watson] and you realize it’s not a good idea to predict too far into the future.
    That 1930s gum you like is going to come back in style!

  13. Back To Top | #273
    trippin' duKeZZ's Avatar
    Join Date
    Dec 2005
    Location
    Bucuresti
    tay is back - imi reinstalez twitter doar pt asta

  14. Back To Top | #274
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    Ar putea fi prea tarziu.

    "Post-lobotomy Tay seemed to have a problem keeping up with the pace of herself."
    "Microsoft seemingly saw this outburst, and it promptly quietened Tay once again and set the AI’s Twitter account to private — meaning that it can’t add new Twitter followers unless they are approved."(techcrunch)

  15. Back To Top | #275
    Spre deosebire de calculator, un om - un copil care invata lucruri in timp ce creste, are si un "drive" natural, care il face sa actioneze.

    Calculatorul nu poate avea asa ceva deoarece nu e viu, daca va avea un "drive", acela va fi programat, si nu vom avea de-a face decat cu un obiect foarte inteligent, dar programat.

    Pe langa lucrurile de baza cum ar fi supravietuirea (care si acelea trebuie programate intr-un robot, nu vin natural), de ce ar cauta un robot - companie de exemplu? De ce s-ar uita uimit la o opera de arta sau la un peisaj, etc, chestii din astea "spirituale"? Nu prea are cum daca nu e programat sa faca asa ceva. Puteti zice ca nu e programat, invata... Invata, dar trebuie sa aiba un motiv pentru care invata, sau pentru care aplica ce-a invatat, altfel nu face nimic. Si motivul asta e programat, nu are cum sa vina din nimic.

    (stiu ca sunt oaia neagra, paria, exilat si excomunicat la divagatii - sper sa pot posta aici, promit sa nu pun linkuri, nici spre CNN)
    Last edited by jet li; 30-03-2016 at 13:35.
    "It is the mark of an educated mind to be able to entertain a thought without accepting it."
    Aristotle

  16. Back To Top | #276
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    S-au gandit deja si la asta. Unii isi fac planuri despre cum ar putea sa le faca "sa sufere" (pe softuri), sa le oblige sa faca ceva ca sa scape de "suferinta". Altii despre cum ar putea sugera placerea. Vor fi dotate (masinile/corpurile asociate) cu toate organele de simt (mai putin gustul despre care se zice ca e mai greu de implementat) ca sa le asigure un schimb de informatie maxim cu mediul inconjurator. Se studiaza si cum ar putea fi implementate "sentimente". Au in studiu concepte matematice pentru toate astea.

    Si oamenii sunt "programati" in principiu. Daca un om e crescut de lupi (de ex.) pana la maturitate, nu mai poate fi "recuperat" (sau foarte greu).

  17. Back To Top | #277
    Intuitive Investigator Benvolio's Avatar
    Join Date
    Nov 2002
    @jet li: ok, fie - cata vreme rezisti sa te abtii sa postezi de-alea...

    Quote Originally Posted by oldshooterman View Post
    De unde asta ca nu putea sa aleaga? Programele astea functioneaza cu deep learning si invata din experienta. Asta inseamna ca "cineva" ar fi putut s-o invete contrariul (pe Tay) si in mod cert au fost useri care au sustinut contrariul in diferitele discutii. Ca urmare a avut de ales in "arabii au facut 9/11" si "Bush did 9/11", de ex. De ce a ales varianta 2?

    Daca mergem pe ideea ca toti de pe Twitter sunt de parere ca "Bush did 9/11", problema e in alta parte.
    Tu nu vrei sa intelegi ca Tay nu intelege. "Ea" nu sesizase ca interlocutorii sustin 2 puncte de vedere opuse in legatura cu chestiunea X; ea doar inmagazinase ce spuneau userii, si cand venea vorba de chestiunea X, cauta in baza ei de texte ce zisesera altii despre X (textele care contineau cuvintele X) si servea mai mult sau mai putin aleator una sau alta dintre opinii (sau poate ultima primita era prima servita). Exemple:

    But while it seems that some of the bad stuff Tay is being told is sinking in, it's not like the bot has a coherent ideology. In the span of 15 hours Tay referred to feminism as a "cult" and a "cancer," as well as noting "gender equality = feminism" and "i love feminism now." Tweeting "Bruce Jenner" at the bot got similar mixed response, ranging from "caitlyn jenner is a hero & is a stunning, beautiful woman!" to the transphobic "caitlyn jenner isn't a real woman yet she won woman of the year?" (Neither of which were phrases Tay had been asked to repeat.)
    Twitter taught Microsoft’s AI chatbot to be a racist asshole in less than a day | The Verge
    "Viata e complexa si are multe aspecte." (©Filantropica)

  18. Back To Top | #278
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    OK, asta e clar (ca ea doar inmagazinase ce spuneau userii), dar din moment ce sunt doua categorii de useri, unii care sustin ca "I love feminism now" si altii ca "feminism is a cancer", de ce a ales varianta 2 in loc de varianta 1? A ales varianta doi pentru ca au fost mai multi useri care au sustinut varianta 2? Nu cred ca au fost o gramada de useri care au folosit exact cuvintele din varianta 2, ca sa repete cuvintele respective.

    La fel a facut si AlphaGo cand a ales mutarea 37, a ales-o din mai multe solutii disponibile, nu era singura varianta viabila in contextul respectiv, dintre toate solutiile posibile. A alege sta la baza procesului de luare a deciziilor.

    Nu e vorba de constienta, dar AI e in faza in care invata sa aleaga solutiile cele mai bune din experienta proprie, ceea ce e o mare realizare.

    "Sometimes, DQN (deep Q-network) discovered game strategies that the researchers hadn't even thought of — for example, in the game "Seaquest," the player controls a submarine and must avoid, collect or destroy objects at different depths. The AI program discovered it could stay alive by simply keeping the submarine just below the surface, the researchers said."(livescience)

  19. Back To Top | #279

  20. Back To Top | #280
    Registered User
    Join Date
    Jul 2007
    Location
    Cluj-Napoca
    Povestea cu Tay mie imi sugereaza o "capcana" destul de bine gandita de expertii de la microsoft. A "lansat-o", a facut senzatie cateva ore, au retras-o ca "sa-i curete creierul", pe chestia asta i-au creeat si mai multa popularitate, a mai scapat putin pe net, au izolat-o pentru un cerc restrans de utilizatori, povestea a atras atentia si deja sunt o gramada care abia asteapta sa povesteasca cu "ea".

    Dar abia au trecut cateva zile si: Now anyone can build their own version of Microsoft's racist, sexist chatbot Tay

    Si un mic amanunt:

    "The data Tay collects is being used to research conversational understanding."
    "Microsoft gathers and stores anonymized data and conversations for up to one year to improve the service."
    "We also may use the data to communicate with you, for example, informing you about your account, security updates and product information. And we use data to help make the ads we show you more relevant to you. However, we do not use what you say in email, chat, video calls or voice mail, or your documents, photos or other personal files to target ads to you."(inverse.com)

    Parca nu e noua povestea asta. (cine nu vrea w10 poate se indragosteste de Tay sau ceva similar)

  1.  
Page 14 of 24 FirstFirst ... 411121314151617 ... LastLast

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •