Silicon Valley Community Foundation - Innovation Conference

Day 2 Session Notes - October 10, 2016

// These are rough and partial notes. Please excuse shorthand and typos. 

Tech training - opportunity or disparity? - plenary

  • Will the push for more coding lead to “technical ghettos?”
  • Black Girls Code is a “pipeline organization” - goal of getting black women into careers. AP CS in high school is benchmark. 5 years old org, first cohort about to go to college.
  • Andela - looking for the best engineers in Africa - “low opportunity doesn’t equal low potential” - 45,000 applicants, 200 acceptances, 2 years old, working next on jobs development side
  • Concern that education system is leaving blacks and Hispanics behind
  • The captains of the tech industry are writing the American culture
  • Current approaches risk leaving these groups at bottom
  • “It’s a continuum” - we have to get from training to education
  • Black Girls Code - some training is highly skills based, but also more holistic, with rooted theory of social change for extended education
  • “I truly believe the college experience is rich”
  • Is it really a ghetto if we’re getting them into solid middle class jobs, given entry level of these wages - e.g. 50k salaries
  • Question about spectrum from training to education
  • Andela - ecosystem requires many aspects, we hope developers will go onto start companies, give back. Our goal isn’t to help them find jobs, it’s to launch careers. We include soft skills like communications, we hope they become tech leaders, 4 years from now look back and say they are also CEOs of international startups
  • NSF - Innovation space exists between the formal and informal education systems
  • 13% of blacks proficient in math in 8th grade in US
  • Make sure the school programs and add-ons are aligned - big opportunity in this intersection, also opportunities to better align with industry / what are the actual jobs
  • “What’s adequate for whom?”
  • If someone is starving, is burnt toast enough? Is it better than what we have, yes, but many people involved here wouldn’t send their own children
  • Black Girls Code - doing first VR workshop next month, staying on top of trends
  • Goal to go beyond where kids already have opportunities, trying to reach those without same levels of access
  • Question about older women
  • Andela - most program demographic is late 20s and early 30s
  • Bobb - we should be explicit it’s about women
  • In 80s, comp sci grads were 40% women, now just 18%, biggest drop in any STEM field
  • Older women are a “forgotten segment”
  • “It’s important we don’t conflate gender and overrun race”
  • Andela - we are 30% women, very active focus to target women and we want to improve these numbers. They didn’t happen naturally, we want it at least 5050, working against huge cultural biases
  • This is an ecosystem problem

Exoskeleton suit - on-stage demo

  • Winner of robotics for good award 
  • Helps people who are disabled
  • Also helps people save energy, which equates to more productivity
  • Targeting 12 industries, including construction and mining

Closing the employment opportunity gap - session

  • Most rapid period of change in history, it’s no wonder we’re trying to figure out the future of work and address opportunity issues
  • LinkedIn - We look at the opportunity gap through data 
  • (I was hoping the session would address ex-felons, but nothing on this was mentioned in opening statements and bios, so I jumped over to the Big Data session.)

Privacy, ethics, and big data - session

  • “Moral rounding up problem” in big data
  • Geolocation discussion
  • Accelerometer data can tell us where you are, even things like if you have Parkinson's
  • Philanthropy can help incentive creation of tools for privacy protection
  • Help reduce bias
  • Tech companies - well-meaning folks, often violate privacy unknowingly, example
  • We need universal human rights for tech
  • Selection bias - pothole discovery using accelerometers, but bias on wealthier neighborhoods
  • Facial recognition - training the algorithms, used college students and celebrities, disservice to other populations like elderly
  • We need more “public interest” technologies
  • We can’t roll back tech, tech is here, policy people are often behind tech
  • We need industry at the table, they’re affecting billions of people
  • Here in US we have “sectoral” privacy system, no federal privacy laws in Internet of things, drones, wearables
  • So policies will probably start at state level
  • Tech tools don’t affect everyone equally, including surveillance
  • Worst affected victim of surveillance was Martin Luther King, everything that happened to him, no one in Senate even mentioned his name when this was being debated (we checked the transcripts)
  • So we need awareness that not everyone is being watched equally
  • There’s a site that lets you search Internet for baby monitors (video feeds), very scary
  • Convening group to look at child online safety, especially around social media 
  • We’re at the beginning stages with all this
  • Takes really long time to develop legislation - by the time you do, the tech changes, so you need set of principles to govern against, so people can govern themselves
  • “I don’t think legislation will keep up with how fast the tech is changing”
  • Initiative around receiving anonymous data from private sector, “magic box”, using it for emergency response like Zika virus forecasting movements, we’re at the beginning of the possibilities these things have
  • Looking for “evidence of impact” cooperation between industry and social justice sector
  • “The [corporate] lobbying you see is utterly overwhelming”
  • A lot of these companies are quite responsible, but what occurs is they hire a bunch of lobbyists, trade associations, consultants - who just go in and stop things, killing bills 
  • Privacy advocates often encouraged to just show the negatives, the risks
  • “In a world where people cannot control their data, they don’t have power”
  • Encourage philanthropy to empower people with their data
  • Question - Where would you deploy $100 million?
  • Making tech investments in underserved communities, we need more engineers of color
  • Bring people together to come up with standards so people can control their data
  • I asked where blockchain fits in
  • It’s well-designed for financial transactions, so that’s what it should be used for. Feels like a tech looking for a problem.
  • I don’t know enough about it
  • Initiatives in smart ledgers and supply chain tracking
  • Summaries - people trying to make measurable impact, funders don’t always want to fund things that need to come before, e.g. the research before implementation
  • Big data isn’t neutral, it’s just as biased as humans, new tech is reinforcing disparities, e.g. stingrays
  • New tech is mostly unregulated with regards to privacy


Online education: the key to equity and diversity? - session

  • “Movement for health of the Internet”
  • “Exponential tech” 
  • “Free world-class education for anyone anywhere”
  • Question - have we seen increases in edu access? More diversity?
  • Khan Academy - still long way to go, online education implies access challenge, recently partnered with College Board (SAT tests), billion dollar cottage industry around SAT prep that favors those with money. Khan Academy created best SAT test prep - free - definitely not a solution, just step in right direction, being disproportionately taken up by students with high needs
  • Singularity U - not just accessibility, also what we teach can be more inclusive
  • Mozilla - online education brings more informal access - the informal system has more equitable coverage, e.g. libraries. Keeping goals in mind.
  • Singularity U - can tech solve this? Only as good as humans can build it
  • Khan Academy - story about classroom how teachers can use tools to better anticipate when student is struggling, pinpoint issues - it’s not about “online education” it’s about education tech that can connect people better
  • Another plug for Weapons of Math Destruction book
  • New “impact fellows” program at Singularity U
  • One University, micro degrees through mobile phones, Kenya smartphone access
  • So how will employers identify these skills?
  • Micro credentials, badges, these standards are being picked up in higher ed, what we need are meaningful credentials that employers trust, it’s a decade long type of problem
  • It’s hard - you need strong signal and recognition. The latter is crucial and under-developed. Who recognizes it and who cares? 
  • Just launching something at Khan Academy to help show equivalent of high school education
  • Biases around how employment is selected feel intractable
  • Learnstorm - a focus on “hustle” points
  • What a lot of employers care about is actually mindset
  • Can this person learn how to learn?
  • A lot of what employers do today will change in the future, so focus on mindset
  • Diversifying social relationships is also really important
  • Open source offers great opportunity where people can come together, learn more social skills while working on same thing
  • Q - if you had $100 million…
  • Story about mayor of San Jose, helping more students get a college education
  • High school credentialing, internationally-recognized, so that it can liberate free global access and help people get into college - can be a transformative solution for underserved communities
  • School grades here have bad logic. The fixed pace for students => ensures variable outcomes => then grades are assigned and you move on. You just proved many students haven’t mastered the previous material. Khan Academy philosophy is to hold fixed the level of understanding, let time be the variable. “Mastery based learning.” Hence building credentials around mastery.
  • Education determines access, and this generation has opportunity to turn education into universal human right.
  • Singularity U - Support breakthroughs, there’s lots of awesome but incremental work, only 5-10% type gains. Example about emotionally-intelligent AI.
  • Social change theory - the great stuff comes from outside the system, not waiting for consensus. 
  • We need to invest in standards, interoperability, and credibility of micro credentials. And the macro credentials like high school equivalency.
  • Q - how to inject credibility into the credentials?
  • All this talk of degrees and credentials can be limiting. GitHub example of showing skills in action. 
  • University degree represents “I did a bunch of work and I met a bunch of people”
  • “Hustle” / “grit” - can actually be tracked - when coming up against difficulty, get help, then breakthrough. LearnStorm’s algorithm actually based on hustle. This helps create more equal opportunities, beyond just rich schools. Oakland won competition, Comcast came in with more Internet access as reward. Positive reinforcement loop.


“Unless we address these issues of growing inequality, the place we call home risks becoming intolerable.” - Emmett Carson


Civil liberties and technology - plenary

  • People in communities of need will use the tools available to make an impact. Story of San Jose mother resisting son’s incarceration using social media
  • Internet is most powerful tool in world for human rights advancement
  • East coast not giving civil liberties issues enough attention, needs to come from within Silicon Valley
  • Concept of Internet access as fundamental human right
  • However there are enormous challenges - can’t be public good without interventions
  • Old fashioned discrimination and invasions of privacy, happening in new ways
  • Facebook - greatest threat is government overreach - where people can’t feel comfortable to share, have freedom of speech, it’s not fair to say US has been leaders of free speech on the Internet
  • ACLU - The tech companies have been very slow to engage in this debate until Snowden, who’s a hero, and then we found they’ve been complicit - they’ve come a long way in a short time, but we have to keep eyes on them
  • Governments are biggest concern - they’re the only ones who can seize assets, lock people up, but still need bigger debate here in SV to ensure public good
  • We couldn’t understand what was happening in SV until we hired computer scientists / experts who know code - we need to engage everyone in this discussion
  • Facebook - agree and disagree - private sector company interests are remarkably similar/aligned with users, since if people aren’t comfortable they won’t use the service. The lack of transparency is the big challenge, we’ve sued government multiple times about, we need to push government for more transparency - absence of transparency has done terrible disservice to this country
  • Silicon Valley Debug - what we see, chief concern, tech brought into police departments, not known until it’s actually used. Suddenly police show up with gadgets looking like they came out of Star Wars.
  • How are people being surveilled by police? We have proven discrimination here, arrests 2x for Hispanics and 5x for Blacks - and now we’re adding more surveillance tools
  • Q about live video
  • Philandro Castile is only known because live video, police just take the phones
  • Not just innovation of tech, it’s innovation of the community in how they use the tools
  • ACLU mobile justice app, now in 21 states - shows what are your rights with police, and intakes info/video for submission to ACLU
  • Its transformed how we work, hundreds of intake videos, we are drowning and we have 1100 staff nationwide
  • Tech companies are not created equal - some have big government contracts, big ones, and hence more likely to sell their customer data, because incentives
  • Also we need to rethink that most laws governing the Internet are from 80s like ECPA
  • We’re precluded from testing / researching algorithm discrimination, due to these outdated laws around computer fraud
  • Don’t even get me started on health records, people can’t even get access to their own medical records
  • We need more rigorous watchdogs 
  • Facebook - and what happens if someone live streams a hostage situation? We’re an even bigger org and we’re overwhelmed too, how do you determine if something damages civil liberties? We’re not sure we’re comfortable with Facebook making these determinations
  • And no laws on that front
  • Q - how are you doing things at policy level?
  • Facebook - not sure first place to go is government, shame on us if we don’t have a privacy policy that users can understand - there needs to be informed consent, there needs to be control. Generally the bigger the company the more controls users get, because there’s more attention on the company. But you can’t go to the NSA and get your data expunged. But you can do that with companies.
  • Q - how do we prepare the next generation on these issues?
  • Our kids and our parents, who also need help
  • SV Debug - We’re always going to be behind if we impose standards of current tech on future tech. We’re stuck in this pattern. So we need to anticipate and have principles - so people know how to navigate. That comes from community, I’m not looking to industry or government for my ethical code
  • Companies seeing us as consumers, governments seeing us as targets or suspects 
  • We can’t relegate ourselves to only respond, e.g. after the new drone or privacy violation - we need communities to lead
  • Q - if a post is deleted, is it really?
  • Facebook - yes, from Facebook, but you can’t make someone forget what’s been said 
  • SV Debug - we see tons of court cases now where evidence is developed based on what’s posted on Facebook - social media and state are intertwined
  • “Nothing to hide” in privacy debate is like “nothing to say” in free speech
  • ACLU - we fought the drone issue - no one was talking about this in America, how we’re dropping bombs in Yemen - but after these technologies creep into our lives…
  • Another theme is AI - if we wait we’ll be too late, so we’re embedding AI people in our org, actively looking for experts 
  • Anecdote about hiring a geneticist at the ACLU, which helped prevent patenting the human genome, in a case that went up to Supreme Court
  • Q - what should philanthropy be doing?
  • SV Debug - Look at the end game, tech as means not ends. This is most vibrant moment of racial justice since we started 14 years ago, in large part because of what we saw on social media
  • FB - 1) access - help underserved get connected, so their voices can be heard 2) encourage people to use the tools - Facebook is free - these tools give you power, it’s an empowering and enfranchising technology 3) fund the experts, those who will keep us honest. But please do so in professional / mature manner, not just to get attention, especially not those who just want to play gotcha. We want people who will roll up their sleeves, we know these are hard problems.
  • ACLU - agreed, 1) emphasizing the access part, ask your nonprofits if they had more money for tech what they would do. 2) more watchdogs, we tend to focus on government but many groups do great work private-sector work 3) we need to force companies to be more responsive to community interests - big challenge is that tech is not diverse, lots of guys. Example about Amazon culture exposé. If I believe tech access is considered fundamental human right, then tech companies better start looking more like all of us.