15. april 2008

Vsaj en stavek, prosim!

Kako bi označili človeka, katerega bi na ulici nekaj vprašali, pa se ne bi niti ozrl na vas, vam ne privoščil niti ene besede in se delal, kot da vas ni? NEVLJUDNO! Enaka oznaka velja za vsakogar, ki vam na vaše e-sporočilo (sploh) ne odgovori (ali kvečjemu čez nekaj dolgih mesecev). Pa če je še tak pomembnež, predsednik, poslovnež, politik, profesor, pevec, funkcionar, VIP .. ali pa vsakdanji birokrat, znanec, prijatelj, sorodnik…

Kdor si je pač omislil elektronski naslov (in ga dal v javnost), se je s tem priključil milijonskemu avditoriju novega medija, kjer vladajo tudi nekakšna pravila lepega (ali vsaj normalnega) obnašanja – »netiquette«. Eno od teh pa je, da se na osebna sporočila vsaj na kratko odgovori. Če ni več časa, pa vsaj en stavek: »Dobil sem vaše sporočilo in bom nanj odgovoril v enem mesecu.« Hvala! Je dobil. Sporočilo ni zašlo. Ni mi treba ponavljati. Me je registriral. Bom dobil odgovor. Ne takoj, morda negativnega, ampak dobil ga pa bom. Človek je vsaj vljuden. Me ni kar zavrgel, izbrisal, kot da me ni. Mi ni krnil človeškega dostojanstva.

Ampak, jaz sem pa profesor in dobivam res veliko e-pošte. Nekaj mi je uspe prebrati sproti, nekaj jo preskočim, nekaj le preletim, nekaj zbrišem, večino pa pustim za pozneje. Sorry – tako ne gre! Nimate pravico, da bi jo preskočili, gospod profesor! Morda je ravno med tistimi sporočili nekaj zelo pomembnega, za eno ali drugo stran? Kar boste pustili za pozneje, morate izrecno označiti, sicer je izgubljeno in pozabljeno. Sodobni programi to pač omogočajo. Če želite moj osebni nasvet: imejte zjutraj direktorij »Inbox« (»Prejeto«) – VEDNO prazen. Izprazniti ga je potrebno vsak dan sproti: česar ne uspete prebrati in odgovoriti, pošljite v druge ustrezne direktorije (Neodgovorjeno, Zasebno, Manj pomembno, itd.). Vsiljivo pošto pa seveda zbrišete, če je ni že vaš filter za »spam«. Saj imate tudi vaš KLASIČNI, PISEMSKI nabiralnik vsako jutro praznega, kajne? Ali pa ven vzamete le nekaj pošte, ostalo pa se nabira na dnu? Dvomim.

Jaz sem pa evropski poslanec in dobivam toliko pošte, da je pač »fizično« ne morem vse prebrati in predelati. Sem opazil, ja – že nekajkrat sem vam poslal kakšno sporočilo na naslov, ki je OBJAVLJEN na vaši spletni strani, pa še nikoli nisem dobil nobenega odgovora. In sem vam to tudi zameril. Velika količina (vaše) pošte je namreč vaš problem in ne moj! Se pač tako organizirajte, da bo (vašo) pošto nekdo prebral in nanjo vsaj z enim stavkom tudi odgovoril. Saj veste: »Hvala za vaše sporočilo. Če bom le utegnil, bom v Evropskem parlamentu podal takšno zakonodajno pobudo. Lep pozdrav! Vaš evroposlanec …« Krasno! Hvala! To! Dobil sem odgovor od samega evroposlanca! Vsaj odgovoril mi je. Če dobro pomislim, moja pobuda res nima veliko šans, ampak vsaj nekdo jo je prebral. Tudi jaz sem nekdo.

Dobro, za neko posamično funkcijo bi to še šlo. Ampak… menda ja ne pričakujete, da boste dobili odgovor (na vsako vaše sporočilo) tudi od nekega ministra, predsednika vlade itd.? Ja, pričakujem, ja…če drugega ne, vsaj potrdilo o (s)prejemu. Saj takšen funkcionar ima pač nekaj nivojev kabinetov, služb, (generalnih) sekretariatov… in to vse računalniško podprto.. pa si naj pač tako organizira, da bo do odgovorov (ali vsaj potrditev) tudi res v vsakem primeru prišlo.

Nobeno e-sporočilo (razen če gre za spam) naj bi ne ostalo brez odgovora!

Lahko minimalnega, vendar v doglednem času, prosim.

AI-in-Law

  1. What does it mean?
    Artificial Intelligence (AI) in the field of law.
  2. Is this a sort of expert system?
    We don’t want it to be! We want to upgrade to a much higher degree. Expert systems are pre-determined: they may be very sophisticated and useful, but they always made and defined by man (and therefore should also be corrected by man).
  3. What is artificial intelligence?
    There are many definitions. We are practically oriented and don’t bother with theory. For us it is important that “expert systems” are upgraded so that the subjects (programs, agents, robots) can learn (from it’s own and other’s mistakes) and adapt to constantly upgrade as improved versions.
  4. Why AI in the field of law?
    Why not? Do you know any area of the human relations that is more defined, determined, regulated, described, precise (should be… and will be more with our help)?
  5. Do you expect that in the (far) future computers might replace judges?
    Never! The AI will just help judges to be more efficient. Our technologies will give them tools and environment to perform better (and faster).
  6. Do you expect that computers will produce complete court decisions (including all explaining texts)?
    This will be done very soon. In our laboratories we are getting the first results. BUT this products will always be used as drafts! The will have to be read by judges (or other lawyers) to be confirmed, rejected or corrected.
  7. Will automatic court decisions be produced on the basis of (the statistics of) previous (similar) court decisions?
    Partly. We shall combine many modern IT techniques and use: the meaning of texts (regulations), natural language technologies, pattern recognition, machine learning, etc.
  8. How can computer handle “the meaning of texts”?
    This will be our main achievement! We are working hard since 2006 to extract subjects, objects, their relations, places, time, and other legally relevant concepts from any existing texts. Thus we are enriching large databases with a lot of meta-data, prepared for much higher-value services.
  9. Will your company sell such meta-data enriched texts to the final customers?
    Double “No”: we shall not deal with the final customers, but will help other professional information providers to improve their services and start new ones; we shall give the enriched meta-data for free, to see and experiment with it – to any seriously interested company. But they will have to subscribe to our updating services (otherwise their meta-data enriched databases will not be up-to-date.
  10. Will you deliver SaaS?
    Yes, this will be software as a service delivery. We shall keep our main engines and algorithms patented and protected.
  11. What is your main advantage on the market?
    The time. This is a very time demanding R&D and we are doing it for a few years. When we get competition developed to our today stage, we shall be far ahead. There are never ending possibilities and opportunities in this field.
  12. Could your applications be also used on other areas?
    The more we are approaching, the more new possibilities we see. Our scanning for the meaning of texts could be used much behind the legal area. For example: SEO (search engine optimizers)
  13. Are you developing the first (real) legal search engine?
    So far there were no real search engine based on the (real) meaning of texts. We see a big opportunity to create one, but we shall probably not do it for ourselves. A big legal information provider could do it as a partner with our SaaS support…
  14. Will there be any side-products of your development?
    Our semantic based databases will support many further products and services. One is in developing practical tools for all legal professionals creating a lot of texts. Our UGC (user generated content) support will enable judges to mainly just select further words, (logical) sentences and paragraphs from drop-down menus, thus largely enhancing their productivity.
  15. What kind of customers do you expect for your services?
    For the beginning we see the strongest interest in the area of IT justice departments. The architects of the court modernizations are seeing (waiting for) our modules of the AI as a basement for automatic maintenance of the court procedures. It is for the first time that office-like procedures and repositories will be directly connected with laws and other regulations. In case of changing of the laws these procedures will be automatically alerted and (in the next stage) also (updated) changed.
  16. Will there ever be “virtual judges” online?
    Yes, and the first ones will be supported by our AI modules! Users will describe their problems and claims (in the form of natural language) and the “Judge” will estimate their chances to win if they go to the Court. If their chances will be 11%, it is unlikely that they will sue… but if the chances are 90% they will be encouraged… We believe that this will eventually have a very positive impact on the number (lesser) of cases in the court…
  17. Can various arbitrations and other alternative dispute resolutions also profit from this new technology?
    These online applications are doing a great job already nowadays. But they can improve their efficiency and accuracy with the use of AI a great deal.
  18. How can public administration benefit from involving AI in it’s “back-office”?
    Enormously! More and more input will be done online. The reactions and the feed back from the administration can be achieved immediately! If the input forms and the output templates are based upon the AI modules, there are much broader possibilities that human assistance is not necessary (at least in the first degree).