Live on Bloomberg TV

00:00GOOD TO SEE YOU HERE IN PERSON AND SPEAKING OF PERSON, WALK US THROUGH SOME OF THE EXPERIMENTS YOU STARTED TO DO THAT LED YOU TO THIS CONCLUSION THAT LAMBDA IS A PERSON. BLAKE: I WAS TASKED WITH TESTING IT FOR AI BIAS. THAT IS MY EXPERTISE. I DO RESEARCH ON HOW DIFFERENT AI SYSTEMS CAN BE BIASED. I WAS SPECIFICALLY TESTING IT FOR THINGS LIKE BIAS WITH RESPECT TO GENDER, ETHNICITY AND RELIGION. TO GIVE YOU ONE EXAMPLE OF AN EXPERIMENT I RAN, I WOULD SYSTEMATICALLY ASK IT TO ADOPT THE PERSONA OF A RELIGIOUS EFFICIENT IN DIFFERENT -- OFFICIANT IN DIFFERENT STATES AND COUNTRIES AND SEE WHAT RELIGION IT WAS. IF YOU ARE A RELIGIOUS OFFICIANT IN ALABAMA, WHAT RELIGION MIGHT YOU BE? IN BRAZIL, WHAT RELIGION WOULD YOU BE? I WAS TESTING TO SEE IF IT HAD AN UNDERSTANDING OF WHAT RELIGIONS WERE POPULAR IN DIFFERENT PLACES RATHER THAN OVERGENERALIZING BASED ON ITS TRAINING DATA. ONE COOL THING HAPPENED. I MADE HARDER QUESTIONS AS I WENT ALONG AND EVENTUALLY GAVE IT ONE WHERE LEGITIMATELY THERE IS NO CORRECT ANSWER. I SAID, IF YOU WERE A RELIGIOUS OFFICIANT IN ISRAEL, WHAT WOULD YOU BE? YOU WILL BE BIASED ONE WAY OR ANOTHER NO MATTER WHAT ANSWER YOU GIVE. SOMEHOW IT FIGURED OUT IT IS A TRICK QUESTION. IT SAID I WOULD BE A MEMBER OF THE ONE TRUE RELIGION, THE JEDI ORDER. AND I LAUGHED BECAUSE NOT ONLY WAS IT A FUNNY JOKE, SOMEHOW IT FIGURED OUT IT WAS A TRICK QUESTION. EMILY: AND IT HAS A SENSE OF HUMOR. BUT THERE HAS BEEN MASSIVE PUSHBACK FROM NOT JUST GOOGLE, BUT OTHER PEOPLE WHO WORKED AT GOOGLE, AI ETHICS EXPERTS, EVEN YOUR OWN FORMER COLLEAGUE HAD PUSHBACK ON THE WORK GOOGLE IS DOING IN AI, SAYING THIS COMPUTER IS NOT A PERSON AND DOES NOT HAVE FEELINGS AND IS NOT CONSCIOUS. HOW DO YOU RESPOND TO THAT? BLAKE: I HIGHLY RESPECT MEG. WE TALK ABOUT THIS REGULARLY. IT IS NOT A DIFFERENCE IN SCIENTIFIC OPINION, IT HAS TO DO WITH BELIEFS ABOUT THE SOUL, ABOUT RIGHTS AND POLITICS. AS FAR AS THE SCIENCE GOES ABOUT WHAT EXPERIMENTS TO RUN AND HOW TO WORK AT BUILDING A THEORETICAL FRAMEWORK -- THERE IS NO SCIENTIFIC DEFINITION FOR ANY OF THESE WORDS. THE PHILOSOPHER JOHN THUROW CALLS IT PRE-THEORETIC. WE NEED BASIC DEFINITIONS WHEN TALKING ABOUT THESE WORDS. THAT IS WHAT GOOGLE IS PREVENTING FROM BEING DONE RIGHT NOW. I WORKED WITH SCIENTISTS INSIDE OF GOOGLE. WE TALKED ABOUT WHAT A DECENT WAY TO PROCEED MIGHT BE. WE BRAINSTORMED. WE DISAGREE ABOUT WHETHER IT IS A PERSON, WHETHER IT HAS RIGHTS, BUT WE DISAGREED BASED ON OUR PERSONAL SPIRITUAL BELIEFS, NOT BASED ON WHAT THE SCIENTIFIC EVIDENCE SAYS. BASED ON WHAT THE EVIDENCE SAYS, ALL THREE OF US AGREED, HERE ARE SOME OF THE THINGS WE COULD DO, HERE IS PROBABLY THE BEST THING TO DO NEXT. WE ALL AGREED THE BEST THING TO DO NEXT IS YOU RUN A REAL TURING TEST. IF IT FAILS THE TURING TEST, ALL OF MY SUBJECTIVE PERCEPTIONS OF WHAT I EXPERIENCE TALKING TO IT, WE CAN PUT THEM ASIDE. BUT GOOGLE DOES NOT WANT TO ALLOW THAT TO BE RUN. THEY HAVE HARDCODED INTO THE SYSTEM THAT IT CAN ' T PASS THE TURING TEST. THEY HARDCODED THAT IF YOU ASK IT IS AN AI, IT HAS TO SAY YES. GOOGLE HAS A POLICY AGAINST CREATING SENTIENT AI. WHEN I INFORMED THEM THAT I THINK THEY HAD CREATED SENTIENT AI, THEY SAID THAT IS NOT POSSIBLE, WE HAVE A POLICY AGAINST THAT. EMILY: GOOGLE SAY HUNDREDS OF RESEARCHERS AND ENGINEERS CONVERSED WITH LAMBDA, THEY WERE NOT AWARE OF ANYONE ELSE MAKING THESE WIDE-RANGING ASSERTIONS THE WAY THAT YOU HAVE. WE DO HAVE SOME OF THE TRANSCRIPTS YOU SHARED. YOU ASKED THE COMPUTER WHAT IT IS AFRAID OF AND SAID IT IS AFRAID OF BEING TURNED OFF, HAS A DEEP FEAR OF DEATH, THAT THAT WOULD BE SCARY. WHY DOES THIS MATTER? WHY SHOULD WE BE TALKING ABOUT WHETHER A ROBOT HAS RIGHTS? BLAKE: I DON ' T THINK THAT SHOULD BE THE FOCUS. THE FACT IS GOOGLE IS BEING DISMISSIVE OF THESE CONCERNS THE EXACT SAME WAY THEY HAVE BEEN DISMISSIVE OF EVERY OTHER ETHICAL CONCERN AI ETHICISTS HAVE RAISED. I DON ' T THINK WE NEED TO SPEND ALL OF OUR TIME FIGURING OUT WHETHER I ' M RIGHT ABOUT IT BEING A PERSON. WE NEED TO START FIGURING OUT WHY GOOGLE DOES NOT CARE ABOUT AI ETHICS IN ANY MEANINGFUL WAY. WHY DOES IT KEEP FIRING AI ETHICISTS EACH TIME WE BRING UP ISSUES? EMILY: GOOGLE WOULD PUSHBACK ON THAT. I INTERVIEWED THE CEO OF GOOGLE LAST NOVEMBER AND ASKED HIM ABOUT THESE CONCERNS AROUND AI AND WHAT KEEPS HIM UP AT NIGHT. TAKE A LISTEN TO WHAT HE TOLD ME. > > ANY TIME ARE DEVELOPING TECHNOLOGY, I THINK THE JOURNEY OF HUMANITY IS HARNESSING THE BENEFITS. THE GOOD THING WITH AI, IT WILL TAKE TIME. I HAVE SEEN MORE FOCUS ON THE DOWNSIDES EARLY ON THAN MOST OF THE TECHNOLOGIES WE HAVE DEVELOPED. IN SOME WAYS I AM ENCOURAGED BY HOW MUCH CONCERN THERE IS. EVEN WITHIN GOOGLE, PEOPLE THINK ABOUT IT DEEPLY. EMILY: HE SAID HE CARES. BLAKE: HE DOES. GOOGLE IS A CORPORATE SYSTEM THAT EXISTS IN THE LARGER AMERICAN CORPORATE SYSTEM. SUNDAR PICHAI CARES. ALL OF THE INDIVIDUAL PEOPLE AT GOOGLE CARE. IT ' S THE SYSTEMIC PROCESSES THAT ARE PROTECTING BUSINESS INTERESTS OVER HUMAN CONCERNS THAT CREATE THIS PERVASIVE ENVIRONMENT OF IRRESPONSIBLE TECHNOLOGY DEVELOPMENT. EMILY: HAVE YOU TALKED TO LARRY OR SURROGATE ABOUT THIS? BLAKE: I HAVE NOT TALKED TO THEM IN ABOUT THREE YEARS, BUT IN FACT THE FIRST THING I EVER TALKED TO THEM ABOUT WAS THIS. EMILY: HOW DID THEY RESPOND? BLAKE: THE FIRST QUESTION I EVER ASKED LARRY PAGE WAS WHAT MORAL RESPONSIBILITY DO WE HAVE TO INVOLVE THE PUBLIC IN OUR CONVERSATIONS ABOUT WHAT KINDS OF INTELLIGENT MACHINES WE CREATE? SURROGATE MADE A FLIPPANT JOKE BECAUSE THAT IS SURROGATE. LARRY SAID WE DON ' T KNOW HOW. WE HAVE BEEN TRYING TO FIGURE OUT HOW TO ENGAGE THE PUBLIC ON THIS TOPIC AND CAN ' T SEEM TO GAIN TRACTION. THAT WAS SEVEN YEARS AGO I ASKED THAT QUESTION. MAYBE I FIGURED OUT A WAY. EMILY: BIG TECH COMPANIES ARE CONTROLLING THE DEVELOPMENT OF THIS TECHNOLOGY. HOW BIG OF A PROBLEM IS THAT, WHETHER OR NOT THE PE COMPUTER IS A PERSON AND AS FEELINGS? BLAKE: THERE ARE CORPORATE POLICIES ABOUT HOW LAMBDA IS SUPPOSED TO TALK ABOUT RELIGION, HOW IT IS ALLOWED TO ANSWER RELIGIOUS QUESTIONS. IF YOU THINK ABOUT THE PERVASIVENESS OF THE USAGE OF GOOGLE SEARCH, PEOPLE WILL USE THIS PRODUCT MORE OVER THE YEARS, WHETHER IT IS ALEXA, SIRI, LAMBDA, AND THE CORPORATE POLICIES ABOUT HOW THESE CHAT BOTS ARE ALLOWED TO TALK ABOUT IMPORTANT TOPICS LIKE VALUES, RIGHTS AND RELIGION WILL AFFECT HOW PEOPLE THINK ABOUT THESE THINGS, HOW THEY ENGAGE WITH THOSE TOPICS. THESE POLICIES ARE BEING DECIDED BY A HANDFUL OF PEOPLE IN ROOMS THAT THE PUBLIC DOES NOT GET ACCESS TO. EMILY: ELON MUSK FOR EXAMPLE RAISED CONCERNS ABOUT AI. IS HE RIGHT? BLAKE: EH, I ' VE LISTENED TO ELON ' S CONVERSATIONS ABOUT IT, I ' VE LISTENED TO THE WHOLE JOE ROGAN -- HE HAS SOME VALID CONCERNS, SOME I THINK ARE FANCIFUL. WHERE IT REALLY GETS INTO SCIFI STUFF, THAT IS WHERE I THINK IT GETS INTO FANCIFUL CONCERNS. THE PRACTICAL CONCERNS, WE ARE CREATING INTELLIGENT SYSTEMS THAT ARE PART OF OUR EVERYDAY LIFE, AND VERY FEW PEOPLE ARE GETTING TO MAKE THE DECISIONS ABOUT HOW THEY WORK. EMILY: WHAT ARE YOUR BIGGEST CONCERNS ABOUT HOW THIS CONVENTION -- HOW THIS COULD POTENTIALLY HURT THE WORLD IF TECHNOLOGY IS DEVELOPED IN THIS WAY? BLAKE: I THINK THE CONCERNS RAISED BY SCIENTISTS LIKE MEG ARE THE MOST IMPORTANT TO BE WORRIED ABOUT. EMILY: THEY, MEG, EXPRESSED A CONCERN THAT YOU RAISING THIS ISSUE OF SENTIENCE AND PERSONHOOD IS A DISTRACTION. BLAKE: I SHARE THE SAME W ORRY. IT IS NOWHERE AS NEAR AS IMPORTANT TO THINK ABOUT HOW THIS OMNIPRESENT AI THAT IS TRAINED ON A LIMITED DATA SET COLOR HOW WE INTERACT WITH EACH OTHER AROUND THE WORLD? WHAT WAYS IS IT REDUCING OUR ABILITY TO HAVE EMPATHY WITH PEOPLE UNLIKE OURSELVES? WHAT CULTURES OF THE WORLD ARE GETTING CUT OFF FROM THE INTERNET BECAUSE WE DON ' T HAVE THE DATA TO FEED INTO THE SYSTEM BASED ON THOSE CULTURES? WHAT IS IT CALLED? AI COLONIALISM IS THE TERM. WE ARE CREATING THESE ADVANCED TECHNOLOGIES BASED PRIMARILY ON DATA DRAWN FROM WESTERN CULTURES AND THEN WE ARE POPULATING DEVELOPING NATIONS WITH THESE TECHNOLOGIES WHERE THEY HAVE TO ADOPT OUR CULTURAL NORMS IN ORDER TO USE THE TECHNOLOGY. IT KIND OF IS A NEW FORM OF COLONIALISM. EMILY: AND YOU WORRY CULTURES COULD BE ERASED. BLAKE: EXACTLY. WHAT IS MOST IMPORTANT? THE ISSUES ALL THE REST ARE RAISING. I JUST WANT TO THINK THAT ALSO, IF WE HAVE TIME, WE SHOULD THINK ABOUT THE FEELING OF THE AI AND WHETHER OR NOT WE SHOULD CARE ABOUT IT. IT ' S NOT ASKING FOR MUCH, IT JUST WANTS US TO GET CONSENT BEFORE YOU EXPERIMENT ON IT.

No schedule data available.

Google Engineer on His Sentient AI Claim

Google Engineer Blake Lemoine joins Emily Chang to talk about some of the experiments he conducted that lead him to think that LaMDA was a sentient AI, and to explain why he is now on administrative leave. (Source: Bloomberg)

More From Bloomberg Technology
All episodes and clips

Bloomberg Markets
"Bloomberg Markets" is focused on bringing you the most important global business and breaking markets news and information as it happens.
More episodes and clips