alisa: i'm alisa parenti in washington, and you are watching "bloomberg technology." let's start with a check of your "first word news." one person is dead and three others wounded after a shooting at freeman high school in washington state. a spokesman says the threat is, quote, "eliminated." the wounded are said to be in stable condition and the school is no longer on lockdown. and electronic frontier foundation have filed suit against the trump administration over searches and seizures of mobile phones and other electronic devices by border agents who, they say, did not have warrants. they say they were in violation of the first amendment and fourth amendment. the u.s. has banned the use of

kaspersky lab software on federal networks. the washington post says they have given a timeline to scrub that russian brand of security software. the u.s. believes kaspersky's kremlin connections make its software a security risk. the death toll in the u.s. after hurricane irma is now at more than 20. a 55-year-old florida resident died after falling off a ladder while preparing his home for the storm. the deaths of now eight patients at a nursing home in hollywood, florida, are being investigated. global news 24 hours a day, powered by more than 2700 journalists and analysts in more than 120 countries. i'm alisa parenti. this is bloomberg. ♪

cory: i'm cory johnson in for emily chang, and this is "bloomberg technology." coming up, waymo gets the green light. u.s. appeals court clears the way. a bloomberg exclusive. special counsel robert mueller's investigation into russian collusion has social media firmly under the microscope. why prosecutors are zeroing in on hackers. another bloomberg exclusive. blue apron ceo talks about the transition from unicorn to publicly traded company to stock market dog. the meal company's strategy for fending off growing competition. a u.s. appeals court has cleared the way for waymo's lawsuit against uber to proceed to trial in october. they claim uber stole secrets for self driving cars. a judge declined uber's request

to send the case to arbitration. anthony levandowski is claimed to have taken thousands of proprietary files from waymo to uber. they also say they tried to scrub the proprietary information. joining me right now is eric newcomer, who covers all things uber for us. i want to backtrack a little bit. basically, this guy worked -- was highly sought after, an engineer, working with waymo, a google thing. eric: anthony levandowski, a super influential engineer, leaves, starts this company called -- cory: and all of us are like, self driving trucks, what's that about? eric: uber buys the company.

cory: paid by the trucking company, but they are not pursuing that at all. eric: uber's focus is, as you know, self driving cars, and they put him in charge of their self driving car effort, then waymo digs around to see if any of their trade secrets have gone with him. that's the basis of this suit. cory: uber wants arbitration why? eric: they don't want it in public. there are lots of embarrassing details that could come out. they were basically saying that waymo should be bound by its agreement to arbitrate things with anthony levandowski, even though they are only suing uber and not levandowski in this case.

cory: they are trying to take the arbitration clause that they had with levandowski. eric: exactly. cory: so, now the trial happens. we will get some discovery. the trial seems real soon. eric: the key of the ruling -- uber sort of suspected that maybe there would be issues, so they did all this diligence and had a cyber forensics firm looking to everything that levandowski had. that report is tightly capped. uber has refused to give it up, didn't even mention it in the beginning. finally, a judge has said, you need to hand that over to waymo. it will be interesting to see what that cyber forensics firm said, because we haven't been able to turn up these 14,000 files. that was the explosive claim at the beginning of the suit that there were all these files that uber had, and so far waymo hasn't been able to say there they are on uber's computers. cory: did they make it from google waymo to the next

employer? eric: right. cory: uber must -- might know it now. eric: uber's argument seems to be we did our best to wall it off. even if this firm has it, which we don't know yet, that was part of our effort to keep it out of uber's hands. cory: what does it mean for the struggles of uber's board right now? hey, this is the former ceo, let's stick this problem on him, or are they going to have to own this? eric: self driving is important to uber. i think it's a key case. it's hard to imagine. anything is possible, honestly. and i don't think it's clear right now. cory: it's a peculiar business model for an asset-light company to pursue an asset-heavy strategy. great stuff, eric newcomer.

waymo's ceo is on stage at the bloomberg sooner than you think conference. here he is speaking with bloomberg's brad stone about when we will all get our own self driving cars. >> i think the answer literally is the name of this conference. it's sooner than you think. we have been working on this at google and now at waymo for over eight years. we have driven well over 3 million miles. we have started to talk a little bit more about some of the simulation models, which are even more important. last year alone, 2.5 billion miles in simulation. to the point now where technology is feeling mature and ready, which is why we are spending a lot of time, in particular in phoenix, but also mountain view -- in something in phoenix called the early rider program, where

we have actual families driving around in our cars and we understand how people in real families would like to use this technology. that's the last part for us, understanding that, as we continue to refine the technology before we are ready to deploy. brad: yesterday was an interesting day in terms of the regulatory framework. the transportation secretary put out a vision strategy statement that called -- which some consumer groups criticized for taking a little bit of a hands-off approach, allowing manufacturers to test their driverless cars on highways. in the same day, the ntsb came out and said highway regulators need to be more active, pointing to the unfortunate, tragic crash of a man last year in a tesla.

where do you see the regulatory environment right now? what are your concerns about taking this tech forward safely? >> that's a great question, brad. if you look at what the administration has done and what we've seen in d.c. the last couple of weeks, it's very encouraging for this technology, because we are in the early stages. we really haven't served our first users yet, so it makes sense that we are careful and flexible so that we don't unnecessarily or inadvertently squelch innovation. what we've seen in the house, which is something pretty special when you think about it -- how many things have the house of representatives united behind recently? i can think of none other than self driving cars. a bill was passed recently that i think is very supportive and in line with what we saw from the secretary, so i think that's great. i think it bears reminding everyone that there is a difference between the problem we are trying to solve, which is fully self driving and removing the human from the car and letting are sensing and technology and our compute do the whole driving task, and the different problem, with driver-assist technologies. there is a difference between those things. brad: that's what the ntsb ruling was on. are drivers getting a little too comfortable with some of these

technologies, taking their eyes off the road? john: it's the fundamental conundrum in this space. we had a pilot experiment where we put some of our employees in some are self-driving cars. we told them they had to be very attentive, that we would be watching them with cameras in the car, and if they didn't behave, we would take this free car away from them. we ended up having to stop that pilot experiment after just a couple of months, because those google employees couldn't stop taking their eyes off the road. they very quickly came to trust the technology too much, and that's the fundamental conundrum of the driver-assist technologies. if at some point the car needs to ask the human to pay attention and you need to take over and the human has fallen asleep, got distracted, is in a very deep conversation with someone, that could be a big problem. it's one of the reasons we

pivoted to a full self-driving solution, where we never call on the human to take over. we are going to do the driving for you. cory: the waymo ceo, john krafcik, on stage at bloomberg's sooner than you think conference. when asked about the lawsuit, he said that uber's behavior was something that waymo couldn't afford to ignore. the saudi's aramco deal, looking to ipo possibly the biggest ipo in the history of world markets. it might wait even into 2019. you can see more on bloomberg.com. we will keep an eye on that story as it crosses. saudi aramco may not be coming to a market near you. the world has had a day to digest the big news out of apple. how does the deal look in retrospect? we will talk in great detail, next. ♪ cory: wall street should be

worried about technology taking their jobs, but maybe they should. according to vikram pandit, who ran citigroup for years. he told bloomberg that develop its -- developments in technology could see some 30% of banking jobs disappear. he pointed to artificial intelligence, robotics, reducing need for staff and roles like black -- like back office functions. apple ushered in a new era with the iphone. the ceo tim cook unveiled a suit e of new products. you probably knew that. everyone has been talking about iphone x and so much more muted stock reaction. what does it mean with a day to reflect, by both journalists, wall street, and the like? the stock is down about 1%. alex webb joins us now. you cover apple in great detail. what's the reaction a day afterwards?

alex: it's sort of business as usual, really. cory: is it the breathless hyping of apple products we all get forced into? alex: part of the breathlessness, the bags under my eyes tell the story of fatigue. we've been reporting about what was to be expected. some leaks in the previous weekend. mark gurman is really good at this stuff. analysts were relieved to see what they expected. there were no big disappointments. there were also no big surprises. some thought there might be a rear facing 3-d scanner. that wasn't the case. perhaps that leaves some space for upgrades in the years to come. you don't want to do too much too soon, then people won't buy the next phone. cory: the self driving phone?

>> -- alex: not there yet. cory: as wall street has started to look at this, some of the financial news, which seemed a pretty good guide for the quarter, maybe a disappointing release date, did not seem to hurt the stock too much. alex: we had reported going back almost a year there were likely to be delays. there are only a handful of factories in the world that make oled. this created some some delays. this was expected going into the quarter. the forecast that apple gave did not necessarily include the sales which will come from the topline iphone. analysts and investors were relieved that other parts of the business, the iphone 8, and things like services, were things -- likely to prop things up in place of the iphone x. cory: a little trivia, oled was created by kodak, and they

failed to benefit from this invention, but managed to license it out. what's interesting and the question remains, kodak, once a great innovator, lost its way. has apple regained some of the notion of apple as an innovator or are these phones being criticized rightly as copycat stuff? alex: there is stuff which other people already have. oled has been in samsung phones for years. the front facing 3-d -- that's innovation. a lot of it is technology off the shelf, and apple has fine-tuned the software element. what's interesting is what apple is doing in the chips. it's the sort of thing which drives the ability to bring software innovations, and therefore they can update ios 11 and 12, but that means old phones become updated.

you and i as the consumer are forced to buy the latest handset in order to get the latest software updates. cory: and they are not handing over the profits of the chipmakers. alex: that's why we are starting to see increasing tension with qualcomm. they cut out a number of suppliers. imagination technologies announced earlier this year they were losing apple as a customer. apple announced its own graphic processing unit, gpu, this chip which is used for games in iphones. the more that apple brings in-house, the more pressure they are putting on their chipmaking suppliers. cory: alex webb, thank you very much. so, coming up next, we are going to take a look at more news coming out of our big conference. things coming sooner than you think. this is bloomberg. ♪ cory: a developing story out of

washington that we are watching

-- social media is now a, quote, "red-hot focus" of special counsel robert mueller's investigation into the 2016 election. possible links to president trump's associates, that is according to u.s. officials familiar with the matter. mueller's team is narrowing in on how russia spread fake, damaging information via social media. it is seeking information from companies like facebook and twitter about information that happened on their networks. facebook said it discovered $100,000 in ad spending, likely from russia, on fake accounts. president trump blocking a chinese-backed company from acquiring lattice semiconductors. the president citing that the security clearance would not happen.

who is lattice? they are based in portland. what kind of chips do they make? >> they are a very small company. $1.5 billion is nothing in market cap for a company like this. they make programmable logic. cory: these are the kind of chips that a company can use, whether it's an automaker, refrigerator maker, cell phone maker, and they don't have to design the whole chip. they can just make the trip do -- chip do certain things for them. >> these chips are used when you design another chip. you map it, know it works. then away we go. cory: they use it because it has a lot of optionality. then you specify a chip once you've done that. >> you can change its function after it has been locked down. cory: it doesn't seem like really complex semiconductor technology.

>> no. and this is a company that is not a large company. the product itself, you can argue, could be used in various ways. it could be useful to the military, but that's not what we are talking about here. we are talking about a policy, thefius change in policy in washington and the hardening of the stance from washington towards china. cory: in other words, if france -- french-backed capital firm or a german-backed capital firm, it might not face the same pressure. >> that would apparently be the case. cory: is there a suggestion that somehow there could be embedded ability to monitor the use of these chips, if a chinese company, government-controlled company could see what was happening with these chips? >> they are used in networking. cory: 25% or 28% of revenue for the company last year. >> a lot of networks have these chips in them that can control

various functions and the updated. i want to stress quite clearly of course security is the stated reason. but, fundamentally, the u.s. semiconductor industry, the u.s. government does not want china coming in and taking those key capabilities away from it. cory: who else competes in this area? >> intel, altera. cory: and the business was a lot more focused on networking and also a lot bigger. >> they divide the market up between them. this is a bit player. cory: there was a suggestion that this was being used in a lot more stuff than it used to be. i thought at the time that the chip was kind of an accelerator into a networking chip. >> this has always been the argument, that these chips have had a specific use case, nearly -- narrowly confined to

communication for a long time, but even more narrowly confined to the design of other chips. the suggestion was that they would spread. microsoft is using them in their data centers now, so that is happening, but not very quickly. it is not so much the market opportunity, but how these chips can be used. a broader point here is we are seeing the chip industry consolidate massively over the last couple of years. we are down to 60% market share for the top 10 companies. essentially, the market has shrunk. china needs to get into that industry. it doesn't have anybody in the top 10. it's going to throw a lot of money at buying its own expertise, buying its way into domesticating the industry. with things like today, how is it going to do that now? cory: other companies get higher prices with the notion that there might be a chinese bidder out there.

>> if you are a company like lattice, what do you do? everybody else is getting together. you haven't been bought, and you can't buy anybody else that would make sense. where do you go? the one haven, with all this money china had said it was going to spend -- that will be our rainy day fund. cory: lattice at home on a saturday night all alone. it's sad. toshiba, speaking of chips, is getting closer to funding its -- selling its chip division. toshiba originally announced a deal with the bain capital, may be ending it at the end of the month. they originally announced a potential $19 billion deal with the bain-led consortium in july. the memo does not prevent toshiba from working with other parties, if they can find one. lots of internal problems at toshiba.

coming up, we will have a sit down with the ceo of blue apron after the company's ipo has face planted. what's next? we will see. this is bloomberg. ♪ >> is 11:29 a.m. in hong kong.

i am rosalind chin with the latest first word news. president trump has blocked a chinese-backed investor from buying a semiconductor, citing national security risks. they are a private equity firm backed by a chinese state-owned asset manager. they are seeking u.s. security clearance, including moneygram's takeover. the pace of china's expansion after further last month factory output, retail sales, and asset investment slowed.

it rose 6% from a year earlier in august. retail sales expanded over 10%. efforts to weigh in expansion appear to be working. a u.s. judge has reversed the veil of former pharmaceutical executive martin shkreli, ordering him jailed immediately after a bizarre facebook host. the judges deemed the 34-year-old poses a danger to offering aty, after bounty for a strand of hillary clinton's hair. he was convicted on three charges. global news 24 hours a day, powered by more than 2700 journalists and analysts in more than 120 countries. this is bloomberg. >> i am juliette saly in singapore, checking markets in asia. it is looking like the rally we have had on the regional index has stalled somewhat. hangan see hong kong's

seng index down 4/10 of 1%. weakness coming through from the industrial players and the consumer-related docs. we do have a miss on the china data. taiwan looking better than it did yesterday, a little rally from apple suppliers. honhail see on high -- under pressure. the aussie dollar still with that 80 u.s. sent handle. -- cent handle. always very volatile. unemployment at 5% percent -- 5.6%. big falls in terms of metal prices. bhp down 2% in sydney. there is upside, tencent in hong kong one of the best performers. samsung electronics up 1%. helping the overall kospi.

will be released in china at the end of the month. that is the state of play across asian markets this thursday. ♪ cory: this is "bloomberg technology." i'm cory johnson in for emily chang. we are continuing our coverage of bloomberg's sooner than you think conference. >> if you think operationally what is involved with getting a system,on home cooking hundreds of farmers for growing ingredients, bringing them in, there is quality control, portioning, packaging, shipping and delivering nationally in a high-quality way. that is incredibly difficult logistics. >> that is the value you are pitching to investors, that you can figure out these hard issues. >> that is what we are doing.

we have a startup of a new center. because it is taking longer to launch it, we're opening two centers in new jersey. it has additional costs associated with it, those are short-term costs. where closing down our jersey city component center and opening up our linden center. we were overly optimistic. we realized that. oh we are working on a number of initiatives both short and medium-term in nature to get linden up and running and take on about half of our network's volume. that is what we are doing. >> you are having to do it all in the public eye. would this have been something that was better worked through had you been a private company? matt: look, you know, you've got to be a confident company and you have to be confident about what we are trying to do. we are very confident in what we are trying to accomplish with

our business. we are building a company that is about meal experiences, about helping people cook at home, and going through a large industry that is in massive transition. as more and more dollars move from off-line to online and as new brands capture share from old brands that are no longer relevant to customers in the world we live in, where people are looking for healthier alternatives, more participatory, emotional brands and experiences. so, we feel really good about our long-term prospects. what does that mean? we have to do the things we need to do to achieve our long-term goals. going public, in my view, is one of those things that has helped make our company stronger through access to capital, through access to the capital markets and public currency for talent and the like. we have to be confident to say, look, we know who we are as a company. and as we execute over time, people get to know us better in the public market and they will appreciate what we are doing.

>> i want to unpack one other thing in that run down that i gave our audience here. the stepping down of your coo. to my understanding, there is a bit of a reorganization of the structure at blue apron. talk us through that briefly. matt: there was an important reorganization we were working on and completed. my cofounder stepped down. as we are moving into this next phase of company life, it is more important than ever for us to be able to innovate fast on emerging consumer needs, understanding our core customers who love to cook and want to cook with blue apron, but they have a diversity of needs. for the first five years of the business' life, we have been very focused on scale, because we saw a very large and attractive market and we felt the need to attract capital and talent and resources to invest in building a world-class brand and supply chain that allows us

to deliver great products at great prices. but now, in this next phase, scale is still important, but personalization and innovation on new products, so we can address new segments of customers, that we have historically not addressed, because we've been focused on scale and monetization and increasingly stronger ways for revenue as a customer is increasingly a focus of ours. we divided up some of the teams more clearly to allow us to go after new product opportunities. we have elevated tim smith and pablo to help with those. >> the timing of it -- this is post-ipo. you pitched your story to public market investors. you turned around, just days, weeks later, and are reorganizing the dna of your company.

why did that decision happen when it did? matt: no company should sit still and do nothing. my job as ceo is to every day do what i can to improve the business, and our job as a company is to every day take one step forward. we do new things every day, and we will continue to do new things. in terms of the exact way that we have personnel structured and the like, we will continue to make decisions when we come to crossroads and see opportunities to improve the company. we are not afraid to make changes. that's our job. >> you talk about confidence and trust, though. some of the investors i've talked to don't like to see these kind of changes so soon after you did pitch a narrative. in thinking back to that, with your stock down 36% since listing, how do you continue to instill trust in the investment community in your strategy? matt: we need to execute and show investors that we are doing

the things that we say we are going to do, and continue to build value by building a great brand, great products, great supply chain, and continue to engage are millions of customers all over the country. and we just have to continue to prove that. people, as they get to know us better, will begin to appreciate the amazing asset that we built and the strategy we are going after. if investors say, we don't want change, i think that's misguided. it's a company's job to continue to change in new market environments, when they see opportunities. the weakest companies are the ones that are afraid to change. i think our willingness to continue to change and evolve and build a business in a gigantic market is something that i, as the largest shareholder of the company, quite frankly, i'm really excited for our prospects. >> you talk about making more money off of your existing users, focusing on that channel, as well as looking at scaling your core product.

those value adds -- talk us through your biggest opportunities there. why are you different? there are so many people out there trying to feed you. why is blue apron different? matt: the food and grocery market is one of the largest markets that exists out there. it's not a winner takes all market. there will be many brands and many companies trying to feed you in different ways. what we stand for is this love of home cooking and the experience of home cooking and the lifestyle around that. our mission as a company is to make incredible home cooking accessible to everyone. the reason we call the company blue apron is because chefs around the world wear blue aprons when they are learning to cook. it's a symbol of lifelong learning and cooking. this experiential component of what we do is core to how we engage our customers. it's important to understand we think of our business as a branded consumer products company, not just a distributor of other people's products. you have to be great on

innovating products around customer needs and have the supply chain to be able to do that in a high-quality, low-cost way. cory: that was at our sooner than you think conference. alibaba executives are dumping shares, they've announced a plan to -- representing 9% of the founder, jack ma, the cofounder, and the ceo. you will hear more in our special coverage. we'll bring you interviews with the founder. will be showcasing those interviews around the world come accommodating in a special half-hour program friday. ♪ >> free is not a viable model, but patience is. >> everybody has a big chance to go digital. ♪ cory: the company is called

sophia genetics. it is a swiss technology company helping hospitals process genome data. 30 million dollars extra to fund expansion in north america. the six-year-old startup uses artificial intelligence and plans to process genomic information. it is planning to use this funding, led by a london-based vc firm, to expand in the medical imaging. sophia works with hundreds of hospitals in 53 countries. mostly europe, middle east, and africa. $59 million has been raised in equity investments since its founding in 2011. the role of artificial intelligence in business was a big focus of bloomberg's sooner

than you think conference. ibm's ceo ginni rometty is reinventing big blue. she is hoping to help these companies ai technology every day as the company's revenues fall, but rometty sat down with megan murphy and was asked about watson and ai's role in health care. >> this is another key point about professional ai. doctors don't want a black-and-white answer, nor does any profession. when you interact with ai, you don't want it to say "here is an answer." what a doctor wants is, give me the possible answers, tell me why you believe it, can i see the research, the evidence, the percent confidence. what more would you like to know? that's really what we are doing. the first cancer took almost a year. we are down to less than 30 days now. by the end of this year, watson will have been trained on what causes 80% of the world's cancers. cory: also at this conference, qualcomm's executive chairman paul jacobs.

he says he is still confident that their nxp acquisition -- he thinks they will still get it at the price that qualcomm is offering, despite calls by some investors to raise that price. >> we are going ahead with the transaction the way that we intended, and that's -- you know, we will see how this plays out. you talk about the regulatory issues as well. so, nothing great in the world gets done as a slamdunk. there are always pieces to the puzzle that you have to work around. this is just another piece of the puzzle, but we feel like it's a fair price. cory: qualcomm has promised that deal will be concluded by the end of the year. it is set to pay $47 billion in its offer. we will go back to the conference. joel weber is on stage right now with david siegel. ♪ cory: turning back to our live

coverage of bloomberg's sooner than you think conference in new york. bloomberg market's joel weber is on stage right now with david siegel, making the case that his company's first and foremost a tech company. take a listen. >> i think that the ideas and the educational programs being developed here will help to prepare new york city for this future that is really rapidly changing everything that we know about the world. joel: so, one of the most unique things about the company is that you are a scientist by , background. you mentioned a couple words already, laboratory being one of them. as a scientist, you are also a financial participant in the

markets. you are an investment firm. talk to me about how that happened. how did you get into that field? david: investing is a really interesting problem from a scientific perspective. a lot of people invest from a very intuitive way -- in an intuitive way, which is totally fine. gut instinct. maybe assisted by some data. but, really, from my view, there's another angle to investing and it actually is truly something where you can deeply apply the scientific method, where every aspect of the process of investing can be converted into some sort of scientific hypothesis that you can test with data. it's very scientific. i have a phd in computer science and artificial intelligence.

it wasn't my intention, my lifelong dream to be an investment manager. i just thought it was a really interesting problem with a lot of data, and i thought that if you look at the problem from a different perspective, you might be able to come up with some very interesting, perhaps even better solutions to the problem. joel: what was your aha moment when you first started doing that? david: i'm not really like that. you know, it's basically all very gradual. when i started two sigma, it was just a continuum. we are working on things. we are looking at numbers. we are trying to figure out a good way to do this. then, eventually, the company got to be pretty big. there was never any one moment where i said, aha, we've really figure this thing out. joel: two sigma started in 2001 here in new york.

so much has changed between now and then. also, over the course of the trajectory of ai. can you talk about what's changed from the beginnings of where you started in the field to what you are seeing now? david: i think that machine learning is really, you know, in the last 15, 16 years has completely taken off. it's really changed the way people think about data. and, you know, machine learning is, in a way, the best method anyone has yet come up with to algorithmically find knowledge in unstructured data. if you have a lot of data, machine learning can be an incredibly powerful tool. and, you know, i don't have to explain to people how much machine learning has changed your everyday lives.

it's impacting pretty much everything that you see and read. machine learning is being used to guide searching on the internet. machine learning guides your social networking. it guides what you are reading on your facebook newsfeed. machine learning is really the story of our times. joel: there's a glass half-full, glass half-empty approach to that. how do you view the world? how do you wrestle with these big questions? ultimately, are you an optimist? david: i'm a total optimist. however, i want to be realistic about change. you can be realistic about the future. i think that ai and machine learning have a really good story. i think that, in the end, it will make our lives much, much better. however, between now and, you know, i don't know, 20 years from now, there may be a very bumpy road. not may. there will be a very bumpy road as society is adjusting to these changes. the question is, how will we

navigate the bumps. if you are driving down a bumpy road and you ignore the potholes, what happens? you get a flat tire. we have to be a little careful to not get too many flat tires along the way. joel: i love that you are making it real for an audience of new yorkers, very familiar with potholes. what are some of those complications that are not being talked about enough? david: machine learning is taking away our privacy. the ability for algorithms to rapidly process data has opened up the possibility of algorithms understanding you better than you may understand yourself. that's something we have to get used to. joel: the implications also in investing -- when you look ahead a couple willwhen you look ahead a couple years down the road, like, how is this going to change the investing landscape? david: machine learning

will basically change how every business operates. and it will basically -- and what it's doing in investing and in other fields is essentially taking the -- certain kinds of work that we thought required lots and lots of human skills and making it possible for a computer to do it better and faster. so, what it's doing is it's destroying certain kinds of skilled jobs. it's allowing us to get better results. machine learning algorithms have been shown to read medical images, in some cases, better than radiologists can. it goes through field after field. it can identify your children in photos probably better than many of you can yourself, which is pretty shocking. so, these changes are going to essentially up the bar, so humans will need to and can compete with these algorithms,

but they will have to change their skills. we will have to change our skills. i'm talking as if i'm not a human. we will have to change our skills to better work in collaboration with these algorithms. joel: for the students coming through here, what advice do you have for them? david: that is a great question. in from my own experience, i will tell you. when i was at m.i.t., i got there in 1983. that was around when the internet was first turned on. i think it was something like 1983. and no one there imagined the impact that the internet would have in such a relatively short amount of time. cory: that was two sigma investment's co-founder david siegel, live from our sooner than you think conference. you have been watching "bloomberg technology." this is bloomberg. ♪

♪ >> the following is a paid advertisement for time life's music collection. ♪ bowser: hey, i am bowser. now, i have bowser's rock 'n roll party. wow, look at this, we have a cool drive-in. we got a cool 1957 chevy, and we have a really cool guy, me. all we need is some rock 'n roll.