cory: tesla reports record sales in the second quarter, why is it still falling short? ♪ cory: i'm cory johnson in for emily chang and this is "bloomberg west." yelp cancels its sales plan and the stock cries for help. one of the highest-ranking women in silicon valley abandons intel after not being named ceo, and the rise of the machine in hip-hop. we will show you a rap battle.

now to the lead, investors give tesla a thumbs up, the company delivering 11,500 model s sedans in the second quarter, above the 10,000 it forecast for itself. still, six months into 2015, tesla is less than 40% of the way to its full-year target of 755,000 cars. joining me now is matt miller bloomberg intelligence analyst kevin tynan. when you analyze these numbers since you are an analyst and everything, how should we look at these numbers? they didn't miss it. kevin: and even if you set them up yourself -- the real factor is going to be the model x suv coming in the second half of the year. and if that's on-time and on target, that's probably a more important metric than what we saw with the first half of the year.

cory: a substantial difference , matt miller, in terms of the growth rate. the number i look at his sequential growth. you will know more about cars than i will forget -- or you will forget more about cars than i will ever know. the year over year thing seems to me, at this stage in the company, the right way to look at it is quarter over quarter. matt: absolutely. and it's the way analysts in the industry are looking at it. you have 15% growth second quarter over first quarter. so even though they sold less than 40% of the cars they have forecast for the full year, they want to sell about 55,000 in the full year and heavily sold 21,000 in the first half, they only have to grow 12% or 13% in each of the consecutive quarters to meet their projections. and let's face it. it's all about the model x right now. if it comes out in this current

or at least in the beginning of the next order, that should put them easily at their target because the demand is huge and the orders have already flowed in for that model. cory: i assume the way these sales rollout is like an oil well, where the most production happens right after the new release. this is a new car they started selling in the fourth quarter, where they had big sales. a little better growth in the third quarter of the cars sale the revamped all-wheel-drive vehicle. is it fair to say that when we see the model x, the demand is high? kevin: yes and you will fill in those orders of those customers who have been waiting in line to get the vehicle. it's not going to be another inexpensive vehicle, and i think when the model three comes out depending on where those units are transacting, it will be the biggest deal in terms of long-term volume growth. very difficult in terms of model

s and model x, which will probably close to $90,000 in -- per unit unless you add options. if you are looking for longer-term volume, it's going to come at lower price points than what we are seeing now. cory: in other words, the $75,000 base for the model s but basically $95,000 is the average cost per car. you are saying the real market is for the model three. when is it expected? matt: the tesla three, by 2017 they should have them in production and be heading out the door. i was going to say in dealers lots, but it doesn't work like that, obviously. i'm not sure how profitable that car is going to be because keep in mind, for the model s which starts at around $70,000 without any options, more than half of

the cost of that car is in the battery. a lot of what elon musk and investors have been banking on for tesla is that battery costs would come down or battery power would double. cory: wait. so if half the cost of a $70,000 car is the battery and half the cost of $70,000 is $35,000. that is the suppose it price of the model three. they can't cut the cost to zero. maybe that can cut it 10% or 20% -- that doesn't get to zero. matt: they hope that battery costs will come down and battery capacity will climb. it just hasn't been happening in a way that silicon valley is used to. it's moore's law of battery capacity. cory: i want to go back to that chart about the timeline of what tesla has said about the shipping data on the model three , because this is important and i don't think people have

focused on it enough. i think -- this is just me -- that the timeline is slipping. in march of 2014, he said it would be 2017. then he said it would be the second half of 2017. then in may, he said it would be late 2017, and then he implied 2018 and the company explained full production won't be until 2018. that sounds to me like they are pushing this out because they're not getting the advances they want. kevin: yes, and to matt's point about the model three being profitable, that's not what is a -- what is going to be important -- it's that they get the product to the market. the market hasn't cared about profitability until this point anyway. it's about eventually hitting the volume targets. i agree. the range is supposed to be 200 miles, so you are not talking about the battery pack you would see in a model s or model x, so there will be savings there as

well. these are high hurdles. the gigafactory is a big deal if you are going to do a vehicle targeted to start at $35,000. matt: i also want to point out if you are an investor or even an analyst watching elon musk you buy now understand the pattern that maybe he doesn't overpromise, but he promises things a lot earlier than they come. you saw in the model s -- they have a standard habit of saying we are going to put a vehicle out this year and then we're going to do it next year, and then it will be in the first half of the year after that. this is normal. you can almost set your watch to see how lazy they are with every product release. cory: we've got to go. matt miller from bloomberg, thank you very much for hanging out with us. and kevin, we appreciate your time as well. yelp is having a change of heart. the company hired goldman sachs

to find a buyer, but they cannot buy one -- cannot find they are one. pulling themselves off the block. the sales process is toast. shares are down quite a bit, down 24% over the year. unlike a lot of services, yelp is profitable -- barely -- but sales growth, though declining it is profitable. coming up, the ceo behind the they have a lot of suitors, but apparently not great ones. coming up, the ceo behind the biggest commercial satellite deal in history. greg weiler will show us his vision of space in the internet. and where hip-hop and big data collide -- we will explain to -- we will explain what twista was doing rapping at bloomberg. ♪

[rapping] ♪

cory: revolving door or glass ceiling? pick your modern architecture metaphor but intel has announced that its highest-ranking woman is stepping down -- she will leave her post as the president of intel at the end of the year. she spent 28 years with the chipmaker, including time as an assistant to the former chief executive, andy grove. she became president in 2013. she's leaving to pursue a ceo job somewhere else. sales of one of the largest commercial rocket deals in history -- one web partnering with richard branson -- virgin

galactic. it's something with the internet beaming satellite to space. this is on top of a massive $5 -- $500 million funding round. dan burton joins us from boston to talk about the company's progress is the company's ceo. also, greg weiler. what are you trying to get done? >> we want to enable affordable internet access for everyone. we are building a system that can provide low-cost access to people in rural areas around the world. cory: are there particular parts of the world that can do that? i can't imagine you are thinking about farmhouses in iowa. >> farmhouses in iowa are certainly one market. globally, there are billions of people that have no access.

the cable and fiber systems only cover a small percentage of the total market. it is super important that we get internet access for everyone. if you are behind and you don't have that access, you are economically disadvantaged from the rest of the world. cory: it is interesting, but is there enough of that market to make a business out of it? talk about why, or at least what the difficulty is in terms of the market at that time. >> think about the telecom industry right now. the third largest telecom operator in the world invested in one web and they have 300 million customers. if you contrast that with at&t or verizon, there's a large market out there not in the u.s., literally billions of people. if you take a company, like what we are doing with the internet of scale -- we are able to provide to aviation, maritime, oil and gas and satellite

-- oil and gas, which are traditional satellite operators. there are 30,000 new commercial airliners coming on in the next 20 years. cory: talk to me about how that works. i've watched the satellite services -- the service is so much better in terms of what you are trying to offer in terms of basic need and connection for planes. i've watched gogo gobble up market share. it's a pain in the box -- butt to rewire airplanes. >> the new airplanes will have fresh wiring in them. the key for airlines is to have a single antenna can work globally and you don't have that. you have one antenna that might work in the u.s. and another that might work in europe. so airlines come as a transfer from country to country, they don't have an easy way to get internet access. cory: and there are a gazillion guest: satellites up there.

-- i need to check with google on that. >> that's not the exact number. cory: what is the exact number? >> 648 satellites could give us the online capacity for the world. cory: is it the same internet access you would get at 35 thousand feet as on the ground? >> you can have 400 megabits per second anywhere in the world, or you can have 50 megabits per second at your farmhouse in iowa. the services that will come over one web are pretty much identical bit for bit for a cable modem. cory: thank you for providing me the opportunity to get out of san francisco. the ceo of one web -- oneweb. do i very much. now is the moment when big data science and hip-hop arrive. elon musk has warned about artificial intelligence. perhaps the rappers should be fearing for their job. check this out.

>> the next big raptor -- ra pper may not be straight out of compton or the lb see, but out of helsinki. they are big data rap fans and scientists. >> i'm a doctoral student in data mining. >> he and his crew created an algorithm to perfect wrap -- rap composition. it acts like google search, scanning a database of songs by 104 different artists and then ranks the rappers by rhyme density. wu-tang clan's number one. nicki minaj, 32, the lonely island, 94. then they pull from the bank to write its own rhymes. he says they are 21% more rhyme -dance -- rhyme dense than raps written by humans. so we at them to the test with a rap with twista who holds the guinness book of world records

for the fastest rap. ♪ >> and now, deep beats. >> show them off on the a you know we ride. i don't know. twista: the lyrics the computer generated are not hot lyrics. the human vibe that comes with listening to music or doing rhymes or singing -- the computer can only go so far with it. >> this is not just about -- there are real commercial uses for deep beat-like software. this is another application of machine learning -- computer code that mimics human thought. machines using this kind of artificial intelligence are expected to grow to a $15 billion market by 2019. >> already now you have your iphone series, which you can

chat with, but we want to make these beats more intelligent and smooth to interact with, you can use these. >> so for now, twista's job is safe from the rise of the machines. twista: the computer has a long way to catch up with twista. cory: a long time to catch up to twist a -- twisteda. -- twista. no i am not going to rhyme for you, that is not about to happen. but coming up, robots rolling into tokyo. find out how these baggage carrying bots could be the beginning of japan's robotic workforce. and the lower of hollywood got the best of federal investigators with the dark web drug marketplace. we'll talk about temptation and the dark web, next. ♪

cory: and now time for the daily byte -- $240,000. that's the amount the lead undercover agent from the silk silk -- from the silk road investigation tried to pocket from a secret movie deal with 20th century fox. the extortions released in san francisco federal court this week. the agent pled guilty to extorting in bitcoins. $100,000 he also pled guilty to using his federal position to seize another $300,000 worth of bitcoin from an unidentified user. in the end, he amassed $457,000 on his undercover crime spree. he's looking at a maximum sentence of 20 years in prison.

and a fine of a quarter million dollars. 20th century fox did not respond to phone calls seeking comment on the deal. with a shrinking and aging workforce, japan is putting robots to work. the tokyo airport is releasing 11 robots to clean the building and carry lunch -- carry luggage, but airport officials say it's only just the beginning. what does it take to really integrate robots into the workforce? this is not a simple task. no one's -- no one knows it more than the clear path robotics cofounder, joining me now from toronto. how hard is it to work with robots? >> it depends on what sort of robots we are talking about. that's why japan is rolling them out slowly, piece by piece. some robots we have been working with for decades in manufacturing, but there has been a lot of media coverage

in the last fears -- last few years talking about robots making it into everyday lives and that's a very interesting question. cory: i went to the amazon facility in central california and they have these robots moving stuff around. they are literally separated by a cage from the humans. it didn't occur to me that the metaphor might be a zoo, but the actual interaction with the robot, they were prevented from having one. ryan: the amazon system they have is interesting. i used to work there a few years ago. that's almost -- it's a very good system, but it is a system that came about 10 years ago and now we are seeing a lot more evolution in what systems are able to do in their interactions

day to day with people. that's much more possible now. we are seeing drones and self -driving cars. all of these systems are much more able to interact with people. that's what many companies are designing for, because the amazon systems are doing a great job for where they are right now. but the consumers are asking for more robotics everywhere. cory: if you think robots are weird, you should see people. how do you deal with the human side of the interaction to get the human ready to work with the robot? ryan: it's a huge challenge. for us, because of that challenge, we are starting in relatively controlled environments, not as controlled as the amazon system, but in manufacturing and logistics and in retail, despite having many pedestrians around, it's much

more controlled than city streets. i know that is a challenge that google, audi, volvo and other car companies working on the self driving cars are facing. how do you challenge the robot sufficiently ahead of time with all of the weird things people will do on a daily basis once they are out on the streets? cory: you are building robots that do all kinds of weird things -- we are looking at pictures right now of robots in all-terrain places. talk about those challenges. those are both removed from controlled environments, but they're also removed from people themselves. ryan: honestly, dealing with people is one of the largest challenges. cory: i find it difficult also. ryan: we are trying to build smarter tools in our own way, tools that can help people so they can focus on what they are good at. that is the real challenge. the robot that is out there in

the middle of nowhere is certainly a challenge, but it is a challenge that has been faced in the past already. to have intelligent systems interacting with people, both as operators as well as people they might happen to come across in their travels, that is the real challenge there. cory: really quick, about 30 seconds -- what is the industry where you think you're going to get the most traction? ryan: we are starting to see tons of traction in manufacturing, logistics, and mining. those have been robot is sized -- roboticized for decades. we are only starting to see increases, but retail is another strong case. it's very controlled, but at the same time, there are tons of stores out there where there's a possible use of robotics. cory: the cofounder of clear path robotics. thank you very much.

the long nightmare is over. emily chang is back next week. thank you for joining me this week. enjoy the long weekend. we will see you on monday. ♪

♪ >> from our studios in new york city, this is "charlie rose." charlie: we begin with our continuing coverage of the greek debt crisis. prime minister alexis tsipras made a speech today, after he -- promising the greek nation that their wages and pensions would be safe. the speech came after he offered a new proposal to the eurozone. he accepted most of the terms from previous negotiations with some conditions attached. joining me now is erik schatzker. he is an anchor and editor in large at bloomberg television. here in new york, peter coy economics editor of bloomberg businessweek magazine. i'm pleased to have both of them here as we try to make sense of