AMA with Ben Livshits

by | Feb 21, 2019 | AMA, Community

Welcome to the thirteenth post in our series of BAT Community-run AMAs.

The ongoing AMA series on Reddit is an eight-month-long event that features various guests from the Brave and BAT teams. The goal of the series is twofold: to give fans of the project an opportunity to interact directly with team members, and to give team members—especially those who operate largely behind the scenes—a chance to share their insights and offer the community a window into their work.

The most recent AMA took place on February 20th and featured Ben Livshits, Brave’s Chief Scientist. Ben fielded both pre-submitted and live questions from Redditors concerning a variety of topics, like how Brave’s machine learning model works, how his academic research background has informed his current role at Brave, and what compelled him to work for the company. Ben spoke about the various ways the team plans to drive large-scale adoption outside the cryptosphere, what he hopes to accomplish in the coming months, and his expectations for SpeedReader—a new approach to the ‘reader mode’ in other browsers—on which he and his team are driving research.

Highlights can be found below, with a link to the full AMA at the bottom of this post.

The next AMA will take place on Wednesday, March 6th, and will feature Marshall Rose, Principal Engineer at Brave.

For the full list of upcoming BAT Community AMAs through April 2019, see below.

bat-chriscat: What are some of the biggest differences you see between researchers and product developers (e.g., in terms of goals, expectations, hopes, etc.)? For instance, how often does something dreamt up by researchers actually land in product, or actually end up feasible for productization?

This is an excellent question! Well, definitely, researchers tend to discover problems differently and once they've done so, they often step back and ponder it longer —although it's my job to shorten the 'pondering time.'

We explicitly aim to work on projects that are novel from the standpoint of research and are directly applicable to our product—the research page has some examples just like that!

It's not to say that it's always smooth sailing, but there's a constant back-and-forth between researchers and other engineers at Brave. This is in part to ensure that our work 'lands' nicely as part of product.

Isndjd: The Brave vision is ambitious, and we don’t know yet if the ad model will see broad user adoption. What makes you think it will, and what is your biggest concern for brave moving forward? What are the biggest roadblocks to a working platform and broad user adoption?

Ambitious, for sure. But we do need to aim higher. 🙂

We have seen privacy-preserving ad models hypothesized and proposed in research literature, so I am excited to help make it come to fruition in a shipping product.

Well, users need to be comfortable with the new ad model and feel that it delivers them some value—either in terms of goods or services advertised or in terms of the BAT rewards they get out of it. If they don't feel that way en masse, we will have a challenge on our hands.

SinfulOath: How are you targeting non-crypto users?

There's something to be said about making the product so easy to use that users don't know that there's crypto underneath. Having said that, Brave has a BAT wallet, making this goal a little harder.

One of the key goals we have is to make it possible to redeem BAT—this way the complexity of wallets and exchanges shouldn't come into this picture for most regular users.

jankfrank: What compelled you to work for Brave?

I have an academic research background—I worked at Microsoft Research for many years and later as a professor.

Brave is in a great number of ways aligned with my existing research interests, which is a rare find. This was a chance for me to take many ideas beyond the research setting.

JulesWinnfielddd: How is the machine learning doing at matching ads?

We build a model of user interests entirely on the client within the browser by observing user interactions with the web. It's key that this model not be allowed to leave the browser, of course, for reasons of privacy. The model has other aspects that are aimed to improve both the quality of the ad match and its appropriateness—we try not to bother the user if they're clearly busy working, etc.

This is a great question—long-term, this is is where a lot of the secret sauce behind Brave's ad matching and delivery model will be.

The goal is to design something that is leading to high-quality relevant ads and achieves more for advertisers with a lot less effort on their part and less annoyance on the part of the user. We believe we can do that. The reason for that is simple—the browser sees all interactions between the user and the web and, therefore, has a very accurate picture of these interactions—more so than, for example, a search engine might.

SinfulOath: What are you hoping to have done within the upcoming months?

There's a lot to be done when it comes to 1) improving the ad subsystem, 2) enhancing ad blocking, 3) making the browser faster... there are many specific projects under this, but these are some of the main themes.

onTheHundt: Thanks for taking the time, Ben! Love the work the team is doing on Brave. Question for ya: Does BAM (Basic Attention Metrics) address issues with previous Pay to Surf companies from the late 90s? (i.e. fraud, specifically bots mimicking use and time spent looking at ads)

We're poignantly aware of the potential for bots and hired yet fraudulent human traffic to compromise the ecosystem and are taking proactive measures to make sure this doesn't happen.

We at least try! Things are not exactly what they used to be. 🙂

alex_the_brave: When it comes to research, do you tend to start with a hypothesis and gather data to prove the fitness of the hypothesis or do you start with the data and attempt to derive insights and conclusions?

As a follow-up, what would you say is the penultimate goal of research in the context of browsing? (Thinking about user satisfaction, security, discovering new means of content delivery, etc)

It's hard to define a single goal—in a growing company there are a number of challenges that emerge almost daily that are 'research-y'—they involve some complex algorithms or just rely on specialized knowledge about something like, for instance, high-performance networking. This is where we come in…

One of our goals is to use the collective knowledge of the broader computer science research community to help the product succeed. We have hundreds of papers we have read pertaining to the projects we work on…

It's both of those scenarios, and more. We try not to do measurements for the sake of measurements and we try not to be empty-handed when it comes to supporting our actions. Lastly, you can't measure everything—sometimes you need to build things and then measure them. 🙂

Kakoulis: How do you see the reaction of the add market to your reward system proposition, are you a rebel? Do they get it?

The night is young, as they say... It's way too early to tell. We need to wait and observe some growth.

One thing to keep in mind is that ad agencies are watching their CTRs—click-through rates—if Brave is doing a lot better than average, we think they'll start paying attention.

YouAreInAComaWakeUp: With Brave being a privacy focused browser, how do you see this extending into other areas of data privacy?

This is a great question, albeit a tricky one! We at the company are huge privacy advocates. We see a strong push for privacy online only getting stronger in the coming years—and blocking third-party ads and preventing tracking the way Brave does is only part of it!

Read the full AMA here.  

Read Joel Reis and Sergey Zhukovsky’s AMA from January 30th, here.

Follow the BAT Community’s Updates here: https://www.reddit.com/r/BATProject/

Upcoming BAT Community AMAs:

March 2019
Marshall Rose, Senior Software Engineer
Catherine Corre, Head of Communications
Holli Bohren, Chief Financial Officer

April 2019
Jimmy Secretan, VP of Services & Operations

Related Articles

Continue reading for news on ad blocking, features, performance, privacy and Basic Attention Token related announcements.

Ad Tech GDPR complaint is extended to four more European regulators

GDPR complaints about Real-Time Bidding (RTB) in the online advertising industry were filed today with Data Protection Authorities in Spain, the Netherlands, Belgium, and Luxembourg. The complaints detail the vast scale of personal data leakage by Google and other major companies in the “Ad Tech” industry.

Ready to Brave the new internet?

Brave is built by a team of privacy focused, performance oriented pioneers of the web. Help us fix browsing together.
Download Brave

Join the Newsletter

Contact

Press

press@brave.com

Business

bizdev@brave.com

Support

community.brave.com

Offices

Brave San Francisco

512 Second St., Floor 2
San Francisco, CA 94107

Brave London

Mindspace Shoreditch
9 Appold St
London, EC2A 2AP

Pages