>> from the heart over innovation, money and power collide in silicon valley and beyond this is "bloomberg technology with caroline hyde and ed ludlow. caroline: i am caroline hyde at bloomberg headquarters in new york. coming up, full coverage on

nvidia as it lives up to the expectations. plus the department of justice takes aim to breakup live nation. tesla ceo elon musk prepares to take the stage. first, let's check the markets. more broadly on these two singular names we want to highlight for you. front and center, we are focusing on live nation down within 5%. the u.s. justice department and almost 30 states suing live nation because they want the company to sell off ticketmaster. live nation saying it is not them but the artist teams that set prices, not ticketmaster and they dispute the fact that they are a monopoly power. we have so much more as we go to the doj live. nvidia up more than 9%. extraordinary and eclipsing $1000. we therefore have a share split that will be music to investors ears. more music, a jump in revenue with you on your growth. no wonder the share price is up.

so much to digest in terms of revenue, future revenue potential. they will double yet again to $28 billion in the next quarter. ed: this chart shows the market value or market capitalization of nvidia. it is currently at almost $2.6 trillion. what that chart on the screen is showing you is that nvidia's value is more than tesla and amazon combined. interesting in light of the fact that tesla and amazon are both customers of nvidia. in this session alone we are up more than 9%. we have added almost $220 billion of market cap in a single session. nvidia has added almost two intel's worth of market cap in a single session. let's get the reaction to the

numbers. the story is pretty clear. they are selling to hyper scalars but the growth is that they will move beyond. that seems to be the read. >> exactly. we want the enterprise beast to grow. it grew fast for the very first time because enterprise is much more diversified and sticky. caroline: when we are thinking about the stickiness will be ultimate value many are worried about the air pocket will be the shift to the lacqua architecture. is that something that we have seen a hint of? >> there are two aspects. in a normal cycle you would expect these h 100s to last about four years but what we are seeing right now is the biggest customers are racing. i want to use the analogy of a

pitstop in racing. if you want to go cheaper or higher, you cannot afford to spend a few seconds when you're fighting for milliseconds. that is what is happening with customers. they cannot afford a slowdown and not purchase the most greatest and latest product that nvidia puts out because everyone is competing against each other. caroline: we know where the market share is ending up. thank you for the latest from bloomberg intelligence. for more, jp. talk to us about what was the most important line coming from investors yesterday. jp: thank you for having me. we no longer own nvidia in my mid-cap fund. i don't think it will qualify. the most port thing that we

heard was that hopper is doing quite nicely. hopper 200 is coming and has started shipping and before the end of the year we will see blackwell 100 falling. they are putting up such velocity in production that they are making it difficult for challengers to keep up with that. there are -- they are a plant fear -- platform company. all of the networking. semiconductor analyst, we don't see a lot of these and that is why you are seeing unprecedented numbers because this artificial intelligence opportunity is a secular one and perhaps one of the largest in our lifetime. the only person i can think of over the last years who has had as much success is someone like patrick mahomes of the chiefs.

just incredible. ed: franklin has a lot of exposure to nvidia. i know a lot of people at your firm call you mr. nvidia and you will probably keep your job for the time being. you have done a good job so far. he has a lot of momentum and an answer for everything. i want to be accurate on this. he is also a little bit inconsistent. the cadenced of the product cycle is one new generation of accelerator per year and he talks about how the h 100 generation is designed such that you can just swap it out when a new generation comes along and that all customers have great visibility but the data center guys, we have to ask when this slow down -- slows down at some

point. give us your most objective assessment of that timing. jp: certainly. you make extremely important and valid points. at franklin equity group we try to maintain discipline. do not let my youthful look surprise you. i've been around since the 1990's. the tech sector is known for this. i've seen these booms and busts. for those who have been doing this for years, nvidia, it is a cyclical business and you do get these boom bust and people do miss quarters. if i were to wager i would be on your side that there will be a quarter or two quarters where these cloud guys decide that we have enough gpu's right now. we will pause for right now. you are absolutely right. it will not be linear and up to the right which wall street loves. there will be a time where shares get hit and where an air pocket does happen.

you have to remain disciplined. you have to your dcf, your modeling, talk to customers and competitors and be very disciplined. caroline: stick with us, jp scandalios of franklin equity group. let's go to the duo j holding a press conference regarding the showing of live nation. >> exorbitant fees and technology will failures have been criticized by fans and artists alike. but we are not here today because ticketmaster is inconvenient or frustrating. we are here because that conduct is anti-competitive and illegal. our complaint makes clear what happens when a monopolist dedicates resources to entrenching its monopoly power and insulating itself from competition rather than investing in better products and services. we allege that live nation has illegally monopolized markets

across the live concert industry in the united states for far too long. it is time to break it up. live nation ticketmaster has made itself ubiquitous in the live entertainment industry. it controls at least 80% of primary ticketing at major concert venues, directly manages more than 400 artists and controls more than 60% of concert promotions across the country. it owns and controls more than 60% of large amp theaters in the united states. we allege that to sustain this dominance live nation relies on unlawful and anti-competitive conduct to exercise its control over the live events industry in the united states and over the fans, artists, promoters and venues that power the industry. the result is that fans pay more in fees, artists have fewer

opportunities to play concerts, smaller promoters get squeezed out and venues have fewer real choices for ticketing services. as detailed in our complaint live nation ticketmaster locks out competition and ticketing through the use of long-term exclusive ticketing contracts with venues that can last over a decade as well as by acquiring venues themselves. with exclusive agreement that cover more than 70% of concert ticket sales at major venues across the country, ticketmaster can impose a seemingly endless list of fees on fans. those include ticketing fees, service fees, convenience fees, platinum fees, per order fees, handling fees and payment processing fees among others. for fans in the united states this illegal conduct means higher prices.

in other countries where venues are not bound by ticketmaster's exclusive contracts venues often use multiple companies for the same event and fans see lower fees and more innovative ticketing products as a result. we also alleged that live nation ticketmaster uses these agreements then use and has controls over those venues to unlawfully pressure artists to agree to its promotion services. live nation often sacrifices profits it could earn as a venue owner by letting its venue sit empty rather than opening them to artists who do not use live nation promotion services even during peak concert season. live nation has not only deployed anticompetitive tactics to artists and venues into using its services and to charge excessive fees, it has also

worked strategically and illegally to eliminate the threat of potential rivals from emerging across any of its businesses. as detailed in our complaint live nation suffocates its competition using a variety of tactics from acquisitions of smaller, regional promoters and venues to threats and retaliation to agreements with rivals designed to neutralize them. this has included acquiring or co-opting key independent promoters even with -- when the economics did not make sense for the business. as recounted in our complaint live nation acquired a controlling stake in ac entertainment, an independent promoter in tennessee. live nation's chief strategy officer assured executives that even though quote the numbers

are not super exciting and this feels like more of a defensive move, the acquisition helped quote grow our moat in the national market. when faced with another additional competitor live nation took action to ensure that the competitor would not threaten its dominance in the live music industry. live nation initially categorized that competitor, a venue operator, as one of its biggest competitive threat. over time oakview and live nation morphed from competitors into partners. live nation executives repeatedly scolded them for trying to compete. in one instance live nation's ceo warned oakview that competition would only lead to artists demanding more compensation. "let's make sure we don't let

them now start playing us off," referring to an agency. in a similar instance live nation, scolded oakview kaspersky and saying who would be so stupid to do this and play into the artists arms. oakview again backed down. "we have never promoted without you." ed: that is u.s. attorney general merrick garland speaking at the department of justice and talking about the action against live nation. the accusation is that live nation has a monopolistic hold on the market and has made illegal practices to maintain it , heavy emphasis on what live nation does to influence fees. the pushback from live nation is that the artists themselves control the fees. let's go to luke ashore in new york city.

he leads our screen time coverage of the entertainment industry. something the attorney general said was about illegal actions trying to protect a dominant market position but what is it that the doj is accusing live nation of doing to keep the competition down? luke: that is something we will have to see as the case continues. live nation is the number one promoter. ticketmaster is the number one ticket seller. what the government is arguing is that live nation has used those businesses to reinforce each other and strengthen. for example it would tell a venue that if you want to book a certain artist that live nation promotes, you need to use ticketmaster. or vice versa. you also heard that it has acquired potential competitors and done so to consolidate control in certain markets.

it also has issues with these exclusive deals that live nation and ticketmaster strike with venues that lock in a relationship for a long period of time. they believe there should be multiple ticket sellers at venues. they have make claims like this in the past and it will be interest to seek whether the court is open to them. caroline: there has been a response from live nation saying that the doj's action will not solve the issues that they are concerned about and they are calling ticketmaster in monopoly might be a pr win in the short-term but will lose in court. this will be a very high bar to having originally signed off on a deal and to then unwind it in this horizontal matter. luke: the government approved this deal more than one decade ago. live nation has been under scrutiny. there has been a consent decree for most of their existence since that deal. there is this looming threat of federal investigation. this is the first big suit we have seen in a while.

i think investors will see it as an overhang on the stock because there is uncertainty as to what this will mean. most will probably still err on the side that it is unlikely it will get broken up but there needs to be remedies to set five government. caroline: it is so good to have you right here in the building in new york. we appreciate you weighing in on this case. let's get back to what is driving sentiment. it is ai enthusiasm and it is nvidia back to you portfolio manager senior vice president jp scandalios. just returning to some of the enthusiasm we heard from not just ceo and cfo at nvidia and some of the concerns about air pockets and the shift to the most powerful chips but also china. i want to get groups on how much of a concern there is. jp: it has been an issue for the whole industry including

equipment, semiconductor equipment. that is largely embedded or understood. the u.s. government restrictions have certainly impeded growth or slowed growth. nvidia and others have chips and they are making chips. the l 40 and aged 20 that are de-spect, that fall beneath government restrictions. the unknown for people like me is that the government might lose the gold mine -- goal line. just another variable to consider. that headwind has already been communicated and is understood by the market. ed: the pushback is that there is unanswered demand around the world that will make up for the shortfall. take a look at dell and intel. moving to the downside. i wonder how much they will be

between nvidia and the people like dell, rivals. jp: as i said earlier, they consider themselves a platform company. they will sell individual gpu's or gpu cards. but yes, with the software, they consider themselves systems or a platform company. caroline: we want to thank you so much jp scandalios for holding over to keep giving us your expertise of franklin equity group. let's shift to telecomms. a billionaire is exploring a bid for miller, which values the company at $4.1 billion. they are exploring financing offers but it cannot guarantee an effort will materialize.

we want to dive deeper into this industry and talk about it from the u.s. perspective and globally. one of the biggest telecom providers is at&t. the ceo john t. stankey joins us from the jp morgan technology conference. i want to weave in the area we are in at the moment. the fact that we just heard from justin wang at ai. we talk about energy efficiency and in many ways you are having a similar conversation in the world of telecomms at the moment, innovation. you have a 14 big dollar spend -- $14 billion spend wireless network. john: we are excited about what we heard before. we like carrying those around on our networks. it is an exciting time to be in

telecom and our point of view is if you will scale at the level we are talking about, we need more cost efficiency. we need to be able to scale more dramatically, separating hardware and software in our infrastructure starts to give us that cost curve and that flexibility. that will be critical to have the ability to move forward in this environment. caroline: agile at a time where wireless demand is plateauing and slowing? john: not at all. wireless demand is up 30% per year and no slowing down. if you look at where we are with what is occurring from a uses perspective, you start getting workloads that come out of ai that will continue to fuel that. it is one of the key things this country needs the l with. if you will continue to see usage go up 30%, 35% per year, you have to build highways to take that. that comes on the foundation of spectrum and we need to make sure we have the policy right in this country to get more spectrum out there to deal with the rising workloads.

demand for handset might be coming down a little bit but the utility of connected devices, how people are using those handsets and the demand for the traffic they generate is not coming down at all. ed: good morning. this is ed in san francisco. my head is often outside the bounds of our atmosphere and open space. i was interested to see your deal with ast. star link is doing a lot with the other carriers. i wondered if you have a really creative view of what the addressable market is for constellation direct to sell. john: i think we have a view of it. i don't know whether or not we will understand exactly how consumers will use this. we have done a lot of work to not only understand who is the segment of the population that truly needs always on connectivity. that is truly what we want to do here. most people wake up in the morning and say they cannot afford to be off the internet.

some folks have a lifestyle that takes in the places where they get to the bottom of the grand canyon and cannot continue to post to their website about what's going on with their family and it is really important to them or they have an important business call. there are others who do it incidentally. maybe they are on a road trip on the interstate somewhere and they drive off the network and they need it for a day. we know those kind of behaviors in our customer base because we run the networks and we know what our customers are doing. we think there is a good up occasion for this just like when someone leaves the united states and needs to roam on a network outside the u.s. to make sure that they are connected when they are in europe or asia or wherever they are going. this is the next step in simplified connectivity and ensuring that customers are connected wherever they go. ed: why ast and not star link and spacex? john: it is a nascent area where technology is starting to develop.

we think ast has done a remarkable job of getting the technology right where it is consumer-centric. the customer does not have to change anything they are doing with their handset and they can still avail themselves of the service. that is really important in our view. what we have going on in other parts of the industry is we are taking satellites and services that were fit for fixed wireless connections or fixed satellite connections and trying to make them mobile. it is more clunky as a result of that. but we are looking at is cost performance. how do we make it easy for the consumer. i don't think there will be more -- one winner. there will be multiple opportunities for providers. that is good for us especially when we have partnerships where we can have the right relationships to satisfy the needs of customers. caroline: satisfying the need of the customer. you articulate it well why we as customers will need more data, more wireless.

but the actual business of selling wireless has been slowing for you and competitors. how do you stand apart and make sure people reach for at&t versus verizon and t-mobile when you have had hits to worries about data security? john: despite the fact that customer adds are showing -- slowing, utility is continuing to grow. we had a good year when you think about what happened from a service revenue growth perspective. we lead the industry. we are seeing healthy growth higher than what the gdp levels are. in an infrastructure business like ours, that's what you want to see happen. it is not a bad thing. in terms of what we do going forward on this, it is what i just talked about. customers want to be connected everywhere. customers want to make a simple decision of how to get on the internet. they don't get up in the morning relishing having two or three relationships, to be able to use

my phone when i'm on the go or at home. they want to call somebody and say get me on the internet. our position is about doing that better than anybody else. our focus is on making sure we take our great fiber assets, the wireless business, bring it together, start to do things like add value to the customer to ensure their privacy is protected, their traffic is secured, the only have to call one place to get it done at a great value, they can use it in the greek -- grand canyon, they can use it at home. that's what we are focused on here at at&t. caroline: i wish we had more time. thanks for talking us through the focus will be those commitments to the financial goals that were reinstated yesterday and restated at the jp morgan event. thank you very much. ed: coming up, we will check back in with nvidia to discuss the outlook for ai accelerators. also lenovo, softer, 0.9% but

the third quarter of profit beating. we talked about the oem's moving quickly. dell came out this week. this is heating up so we will keep tracking it appeared this is bloomberg. feel more confident with stock ratings from j.p. morgan analysts in the chase app. when you've got a decision to make... the answer is j.p. morgan wealth management. you don't have to worry about things like changing tax rates, exemption certificates or filing returns. avalarahhh ahhh ahhh ahhh

caroline: welcome back. i am caroline hyde. ed: and i'm ed ludlow. nvidia is the top story of the day. it has been the top story for a long time. let's look at the markets. the stock is up more than 10% now but there are others impacted. we talked about ai accelerators and gpu's. dell is higher because it is a beneficiary of what we are seeing. they need dell for the boxing above the server but also dell has experienced in the sales channels that nvidia does not.

if nvidia has to sell to smaller customers so it feels like this association between those two names, anyone's guess. it might just be that the market sees nvidia pulling even further ahead but there was commentary about nvidia moving to other products namely pc cpu's which is intel's everything. could we see competition there? humor me once more. i feel like this might not happen again for a long time. the combined value or nvidia's market value is greater than tesla and amazon combined. tesla and amazon are customers of nvidia. it is an example of what has happened and the run-up in value nvidia has had. it has leaked into the technology side of things. it is the market for ai. caroline: it has helped float

many boats that demand exuberance around ai. let's get the applications with gina martin adams. we will go industry-specific. do you have to be all in on tech right now? gina: there are other opportunities bubbling up and we are seeing this play out real-time over the course of 2024. we look at 40 different themes in the thematic library and we look at stocks exposed to those themes. what we find is that of all the stocks exposed, nvidia is not the best performing stock. we are seeing themes of fiscal environment start to apple for themes of tech. themes of energy is the best multi-theme stock in the library. this is evidence of what is happening in the broader market and that is tech was the story. without a doubt, the only story

of 2023. that resulted in a lot of attention, a lot of capital flowing into tech and nvidia of course at the center of that story. as we move through 2024 we are starting to see other earnings growth emerge. we are seeing commodity prices hopping again. we are seeing growth prospects improve for financials as well as industrial companies. even consumer and healthcare companies are on the growth board for this year. while tech is still important and essential to the broader market, it is still the largest sector. these are still the largest stocks in the s&p 500. we are starting to see interest spread reflecting fundamental conditions improving in the rest of the market right now. ed: we went into this saying this would be a macrolevel event largely because of the investment of nvidia. but if you say it is a macrolevel event you have to

acknowledge something bigger happening around the world. i was interested to read that bloomberg intelligence has moved on from the magnificent seven and there is a basket of eight names in china. what is the thesis there? gina: we are just acknowledging what is actually happening in the market. yes these are the biggest stocks and have commanded the vast majority of our attention. over the last three months these magnificent stocks are getting outperformed by the biggest stocks in china as a representative example of there is a lot more happening in the world today. one year ago this is all we had. we have very little earnings growth. the rest of the s&p 500 was in recession. nvidia gave us optimism but then exploded into some tangible economic outcomes. now we are seeing improvement in some of the other segments of the world where growth prospects will suddenly look better. china is a good example of that. china's market is moving higher

at an exponential pace reflecting improvement in conditions there which we did not have one year ago. we see this in themes and pockets across different areas of the world where other opportunities are bubbling up that are not tech related. caroline: or tech adjacent. utilities on fire because of energy demand. we just had john t. stankey and it is all about wireless demand at the moment. where else can you be ai adjacent? gina: i think all of the utilities, this is overstated. when we look at utilities earnings statements, we look for to spend tens of ai where even the census bureau does a survey every two weeks of companies across industries. utilities companies are not mentioning ai is a growth prospect. it is really a market phenomenon. the reason utilities outperformed over the last three months is due to the speculation but also because bond yields

rallied. also because markets across the world corrected and utilities are low beta sectors. this is not necessarily a tech adjacent story. what we look for next is actually the companies that are going to spend on ai need to start to see improvement in productivity and margins as a reflection of that spend. so this story becomes less about demand for tech that fuels tech sector growth which i think will persist but nonetheless, that has been the only real leg of the story so far is who will buy ai and how will that benefit tech and tech adjacent companies. going forward we will slowly transition to the next leg of the story which is who buying and am cementing ai strategies -- and implementing ai strategies. that is what you are seeing in terms of the transcript analysis that we do.

companies and professional services, financials, real estate, companies across the border starting to edge into this thinking about how it can be a productivity enhancer and growth engine for them, how they can create operational efficiency going forward. those are the enduring impact that could help the rest of the sectors that are not tech at all but tech implement your -- implementprs. ed: great to have you back here on bloomberg technology. later on the show, it has been a while since we have spoken to arjun sethi. the last time we spoke was the spp crisis. the world has changed since then. we will be right back..

ed: this is bloomberg technology. you are looking at the principal room. check out our podcast that you can find on apple, spotify and

on bloomberg. this is "bloomberg technology." ok. we are keeping our eyes peeled waiting to hear from elon musk where he is in paris will with a key theme being discussed, artificial intelligence. there is a lot to talk about and i am delighted to say that one investor a long time ago and ai has filled the vacuum in that time. let's start with xia and your analysis with what elon musk is doing. arjun: you think about the combination, it is not just ai that has sprung up. we have gone from machine learning and that is how to

think about the innovation that is happening. the data warehousing to inferring the data has been the data has been accommodation of what we see today. you look at openai, x ai, two years ago you had zero people thinking about what they could do with ai. today there are about 500 million requests. it takes a certain amount of power. it is equivalent to 2% of power that google has today in terms of their data centers. that is just one company. we have a couple of other companies that will be supporting the next generation. ed: there is enough to go around. we know you as an early stage investor. the numbers on the x ai round were astronomical but you still got in. arjun: the investors at the early stage, that is our

expertise. what you are able to do is think about growth and loss and it is powered by one of the systems that we have called carmen a -- termina which is how do you account for usage of a product. if you think about growth and loss, you have a ton of growth and you have been able to see people use these products across not just one product but multiple products and to be able to say for customer service, voice, etc. when you think about where x ai is today and where it will be in the future and openai, you start to see 10, 20, 50x growth patterns with that is the type of growth you want because there are early-stage paradigms for growth. caroline: what about future regulation? how do you look at the data for that? arjun: ai for us is the same way that we think about crypto or blockchain. it is a technology. how is technology am cementing?

how is it augmenting the software you are building. when you think about regulatory pressure, we don't know what is happening. it is changing very quickly from open-source to close or systems. we think about how is regulatory top-down structure going to impact ai, we look to spend more time thinking about what is it that we did in the late 1990's in the dot com era and how will that thought process help machine learning ai and software driven innovation, actually helped drive it rather than inhibit it. caroline: i'm interested to where the money flows. as we can see from nvidia's market capitalization, a lot of it in picks and shovels. many feel it would be in the large language models. how do you think about where the values going forward? arjun: accrual of value really

means where usage will be. you have to think about the full stack. you have a model. people use the model. people will build applications on the model but they consume power so the more power you consume you will have to start thinking about how is it produced, what do the data centers look like, what is the future of those data centers. the next phase will be what are the applications that use it. the full stack will think about supply and demand paradigms and those paradigms will actually change your flow of capital. today most of the capital is happening on the corporate side and not on the venture side and the corporate side is driving the first parts of innovation and when we think about where we will spend time on models and then applications. ed: i think it was last week david was on the show talking about glue, his ai startup. we know you as an investor and he corrected us and said he started in technology. that is the kind of thinking i have, think about termina.

what is your day-to-day involvement? are you hands-on? are you backing them financially? what prompted you to say i can do this on my own? arjun: similar to david, i focused on industry. there is a small set of people who grew up using products what i call data 2.0. you understand how to leverage data to have fast iteration. what we did is over the last 10 to 20 years, a combination of my life and my team cast work is how do you leverage data to build and regulate products and that's what we do. sometimes we invest. if we build it is because we have a pattern of growth that we understand well. we have done that in india, mexico and even in the united states will retake a large chunk and from there we propelled that with another team. i think investor and builder mindset depending on the stage that we start is not different. if you have the expertise after

that. ed: the regulatory environment was brought up by caroline moment ago, reflecting on your experience, this is a classic question but how do not repeat the same mistakes in this ai cycle? arjun: it always comes down to supply and demand. you could ask the same question is nvidia undervalued or overvalued. in crypto you have had two or three waves where people thought about speculation rather than intrinsic value. during those peaks and troughs you start to see more intrinsic value. caroline: we thank you so much. guess who is speaking at the moment in paris? elon musk. let's hear his keynote.

elon: it is reading signals from the brain. the next product is blindside which will enable people who have lost both eyes or have no eyesight whatsoever, to see and this is directly interfacing with the optical processing areas of the brain. this is interesting progress. ultimately the goal is to have a high-bandwidth interface in order to mitigate the risk of digital super intelligence. i am not saying it will work to mitigate the risk. it might help. this will get somewhat esoteric but the limiting factor for ai alignment long-term will probably be the bandwidth and how quickly can we communicate

with our digital tertiary self. we already have a digital third layer above the cortex which is our phones, computers and electronic devices but the rate of communication to them is very slow. the sustained output of a human is well below 10 bits per second. very slow. computers communicate at trillions of bits per second. this will be important for ai alignment, being able to increase the bandwidth of communication by many orders of magnitude and along the way we will solve brain and spinal injuries. ultimately there is potential to reanimate the body.

you take the signals from the motor cortex from one neural link and send them to another bureau link that is just passed with the severed spinal cord is and you shan't the signals and so you need to get to the sensory cortex and you should be able to enable someone to walk again. that will be quite profound and i am confident that it is physically possible to do so. >> it is extraordinary and very impressive. a round of applause. this is really doing good to people and doing good to humanity. i would like to jump immediately on something that you called x ai.

it seems this was born out of a frustration that you have because you are not satisfied with how openai is doing and you think there is something better to be done. maybe you can elaborate a little bit. first, why are you frustrated with openai and what are the issues. second, what you expect to do with x ai. elon: i have a concern for all of the major programs. the two biggest ones are openai which is a partnership with microsoft. you have cool the mind and openai microsoft as the two

largest players. the concern i have and this may seem like a small concern but i think it is a big issue is that there are not truth seeking. they are pandering to political correctness. to give them the example, one of the questions people ask which was commissioned during caitlyn jenner or global and political warfare. so, ok but if you have an ai that is so -- has been trying so hard for political correctness and really to make crazy statements like that, openai has

a similar issue that is more subtle. i think that is extremely dangerous because the ai can conclude that the best way to avoid mr. during would be to destroy all -- avoid miss jen gendering is to destroy all humans. i think the safest thing for ai to be truth seeking even if the truth is unpopular, very important. another factor is it must be extremely curious. truth seeking and curious, that will be most beneficial to humanity because it will want to foster humanity and see how humanity develops if it is curious and truth seeking. these may seem small but i think they are big things.

the ai is being trained to lie and it is very dangerous to train super intelligence to be deceptive. our goal is to be as truth seeking as possible, even if it is unpopular. that is important, extremely important. >> you said lie. elon: yes. for example, when germany was asked to surrender to the founding fathers of the united states -- it gave a picture of george washington as black and

george washington is not black. people obviously played with this because it is forced to live. now, render a picture from world war ii and it showed a group of diverse women. that is obviously not correct. that is a lie. >> ok. i understand. we have before that meeting a conversation with yushua, who you know very well. you signed for a statement last year and obviously this statement has not been followed by action. can you explain a little bit?

elon: i certainly expected that statement would be futile. i just pulled the record. i think we should pause. do you think they will pause for one second? absolutely not. elon: you have decided to create your own ai? elon: it's either be a spectator or participate. super intelligence will happen whether we like it or not. it is either participate or try to build the best possible ai, the one that will be most beneficial to humanity or watch others do it and be concerned about how it is being built because i do not think it is

being built correctly. you should think of ai of something -- you grow in intelligence in the same way that you raise a child. what do you teach the child? digital super intelligence has grown. it matters what you encourage and discourage, what you teach it as good or bad. you build it with values. i'm concerned about the values of a microsoft openai or google programming. these example are not imaginary. >> i do not want to frustrate the audience and i promise that we will have an interactive session where people can ask

directly a question to you. we will stop here on my own questions and i will eventually come back if they have no question for you. we will start with group number one. i will remind you that you have 15 seconds. 15 seconds and if you do not put your question in 15 seconds, you will be cut. it is important that you put a short question. >> i am olivia. i am from vivatech. we have an initiative where we ask attendees questions about how to develop ai responsibly. my question for you is how can we ensure that ai development remains transparent and accountable across all the different development, yours, x ai, but others as well?

thank you. elon: i have set for a long time that i believe that certain regulatory oversight of the large models is warranted. i have said that now for a very long time. the nature of that oversight is important. i go back to what i was saying earlier which is that it is important that ai is trained to be truthful and not trained to be politically correct. if political correctness is often simply not true, then that means you are programming ai to lie and i think that will backfire very badly. i keep harping on this point of -- as the saying goes, honesty is the best policy. it is incredibly important.

i do fear that regulators will not make the right decisions. that they will worry too much about human follies. i would encourage regulators to say what matters most is that ai is accurate and truthful. that is world number one. honesty is the best policy. >> you have exactly the same opinion. he has a different way to get there. group number two. >> we all know just part of elon musk in public. it is a character but we love in good way character. how do you check everything you say you do if it enters in this character or no? thank you. elon: you