This week I attended the two day Gigaom AI conference here in San Francisco. For those of you unfamiliar with Gigaom, they are the industry leader in emerging technology research. The focus of the conference was on how companies are innovating with AI to drive ROI across different aspects of business, from industries like home improvement and legal, to sales, marketing, and customer experience. I was surrounded by top enterprise executives, investors, and entrepreneurs joining Gigaom’s research customers in a beautiful, high energy setting. Over 25 speakers on cognitive software, workflow automation, virtual assistance, and more. You can follow along the hashtag #GigaomAI conversation here.
Here are a few high level takeaways:
1 – AI is not only coming, it’s here, and if you’re not investing time to learn about it now then you’re falling behind.
2 – AI is not about the sexy products and services as much as the backend, behind-the-scenes functionality.
3 – AI will be as big, if not bigger, than both Mobile and the Internet.
I was personally blown away by the content, use cases, and forward thinking attendees. I couldn’t help but to think about Singularity University throughout the sessions. If you’re not familiar, I highly recommend you check it out.
So I volunteered the entirety of day one, but caught the afternoon pitch challenge for AI startups fighting for an investment prize and focused on serving corporate customers. Here’s a quick rundown of the judges and each startup concept.
Investors and Judges
- Kimberly Powell – Sr. Director, Deep Learning & AI, NVIDIA, introduced the NVIDIA Inception Program
- Bill Ericson – Founding Partner, Wildcat Venture Partners
- Rudina Seseri – Founder and Managing Partner, Glasswing Ventures
- Howard Love – CEO, LoveToKnow, Angel Investor
First up: Netra
- Represented by Richard Lee, Boston-based CEO of Netra.
- Netra allows you to understand how consumers are engaging with your brand or category, and improve targeting using their Visual Intelligence solution.
- A really neat demo — Making sense of consumer imagery by adding structure to better understand decisions and preferences, on a corporate level. You are not your social profile/posts (a self-built report), more so a culmination of the pictures you take. Tells a lot more about you: golf trips, family vacations, hobbies, locations, home renovations, food, etc. etc.
- Consumers post over 3.5 billion photos on social networks every day, and businesses are creating more visual content than ever. Each image posted (or engaged with) provides insights & clues into the consumer psyche – beyond what is self-reported in a profile or captioned in the photo.
- Represented by Matt Peters and Brennan White, cofounders at Cortex and Pandemic Labs. Boston-based.
- Cortex contributes to your brand’s social content strategy. This one really resonated with me, coming from a social management and content marketing background. Cortex allows you to save time with social through content automation and AI. Better decisions, the first time around. Cortex works with the State of Utah and other large clients.
- Create engaging content that inspires action. Outperforms industry benchmarks by 300%! Machine learning under covers customer patterns that make the consumers engage and buy. Content guidance, competitor tracking, and intelligent distribution.
Third: Legal Robot
- Represented by Dan Rubins, Co-Founder and CEO. SF-based.
- Legal Robot allows the every day consumer to operate in correct legal language, automate legal processes, and understand intelligent legal analytics.
- Legal Robot enhances (or replaces) traditional legal processes like contract reviews with an automated intelligent assistant. Suggests improvements by considering best practices, risk factors, and jurisdictional differences. A very necessary solution, seeing as millions of dollars are wasted on a company’s bottomline each year due to contract errors or misunderstandings.
- (PS – they have some pretty awesome swag)
Next up: Kylie
- Represented by Founder & CEO, Jamasen Rodriguez. SF-based.
- Marketed simply as an AI that can “help draft responses to your emails,” Kylie clones employee personalities to automate conversations between a company and its customers. Utilized on an enterprise level across email, support tickets, chat apps, and social media communication.
- 24/7 email monitoring and the option to send automated responses. Rule Your Inbox and start a free trial today.
Final pitch & pitch WINNERS: LeadCrunch
- Represented by Olin Hyde, Founder & CEO, San Diego-based.
- Solving the multi-billion dollar lead generation gap, LeadCrunch is a Customer Discovery Platform that uses artificial intelligence to analyze your customer data and help you target the right contact at the right company at the right time with the right message.
- LeadCrunch provides accurate and valuable B2B leads as the first self-service AI B2B lead generation tools specifically designed for SMB’s.
- Companies like Verizon have given LeadCrunch access to over 60K client profiles, the largest data set LeadCrunch is working with, to help identify new targets.
- Here’s how it works:
More on SafeGraph here: http://www.safegraph.com/
Day One Keynote: Sandy Carter
I unfortunately also missed this one, volunteering, but found some awesome tweets. It was my pleasure to meet Sandy and chat a bit about the conference. She’s a true entrepreneur and leader in all things AI, IoT, and VR. Sandy was the driving force of Innovation at IBM for the last decade, and author of three books including Get Bold: Using Social Media to Create a New Type of Social Business. More on Sandy here: http://sandycarter.net/
Day Two Keynote: Lowe’s Innovation Lab
Kyle Nel – Founder and Director of Lowe’s Innovation Labs, Lowe’s Companies, Inc.
Kyle’s Goal: Help people make better decisions. The Innovation Lab is about exponential change in the future of retail. There’s some really powerful mission statements, goals, and accomplishments on the website – and Lowe’s is a top success story out of Singularity University partnering with Made In Space to create 3D printers for the International Space Station. You’ll want to follow along, for sure.
Some notes from his Keynote:
On Innovation … after a while, the innovative people start fading out of “innovative” companies. Gradually improving a product over and over and over is not success, it’s a Linear Path to Doom (and TPS Reports).
We live in an exponential world. [Tech and new concepts] can be deceptive at the beginning (AI, AR, VR), when the hype doesn’t match the expectation, so people end up ridiculing it. It’s hard to see how new concepts will scale, etc., and when the hype wanes, only the core group of people will continue working on it (the “deceptive disappointment” part of Kyle’s curve chart) . To make expectation meet reality requires massive disruptive opportunity.
Chick-Fil-A shoutout: Coming from Atlanta, this is 100% noteworthy. Kyle mentioned Google Cardboards with a cow on it. Eat-More-Chicken, something something. He loves Chick-Fil-A (good, we can still be friends, Kyle), but why? I agree. Perhaps it’s juuust a little unnecessary, and to me feels forced in a lot of ways. But hey – if it’s a kids first time playing with VR, and it’s a memorable/meaningful experience, it’s a win for Chick-Fil-A.
On The Drake Equation … There are ~ 160 Billion planets in the galaxy, so the probability of discovering life is incredibly high.
The Fermi Paradox … Well, where is everybody? This is how Kyle feels when he talks to companies about applying innovation and AI. They spend money tinkering around the edges but don’t have the enormous investment or success stories. This space is a paradox, in a way, that requires serious time, talent, and capital.
On Opportunity Overkill … Say you’re in a new position at a new company, people are acting in a weird, new way. The culture is off, but you adapt to it. The company encourages practicing choice, and blue sky ideas. Here comes the Paradox of Choice, which actually leads to doing a whole lot of nothing. People demand curated choices, and often short circuit in situations of “vast opportunity.” On the topic, Kyle recommends Dan Ariely’s Predictably Irrational book.
On Conspiracy Theories … Sometimes conspiracy theories are logical, but also considered unorthodox and bizarre. If you get stuck on one little fact, you are all of a sudden the crazy lunatic conspirator. Take the JFK Assassination, Nov. 22, 1963 – Did anyone see it coming? Who shot him? Why are people still asking this question? Critical lost hours were spent looking at a weird, superfluous umbrella man because of an anomaly. Takeaway: be sure to invest the majority of your time where it’s most impactful and systematically reconsider the direction you’re heading.
On Newton’s First Law … Object in motion stays in motion… Including organizations. People don’t like trends and change, and companies will continue down their traditional and irrational paths. People don’t want to change. Look at the Google Trends around gym membership turnover between January and March each year. Take this into account when raising new ideas and building new concept teams.
On Brainstorming … His Lowe’s Innovation team brainstorms by reading and fantasizing comic books. “Ooh, how are we gonna do XX??” If Lowe’s isn’t thinking ahead in AI and robotics, it simply won’t be a part of the future.
The LoweBot … Kyle and his team ended up working with a startup to build the first fully autonomous robots (dubbed the LoweBot), currently in use at Northern California Lowe’s stores. It speaks 27 languages, performs inventory store tracking as it helps customers, collects information and conversation trends. Things that have never before been tracked, all while simultaneously performing additional chores. Shows people ads for new products based on their character. Works with hearing and visually impaired customers. Oh, and it runs on a hand-built AI system to get smarter. Boom.
Made In Space … Lowe’s Labs recently collaborated on 3d printers for space, specifically the International Space Station, opening the door for on-demand tool creation.
Holoroom … A virtual reality design experience showing how to use VR in a real environment. Have worked with the Microsoft HoloLens, Oculus, and now an ongoing collaboration with Google Cardboard vending machines in stores (to create your own personalized Cardboard viewer) – very cool. You can hit up the Lowe’s in Seattle for a HoloLens demo today. Future of Retail 101.
Pinterest-based Remodels … Lowe’s is working with Pinterest and Microsoft to envision what a remodel project of the future could look like. It’s a fascinating process that people put themselves through, seemingly unquestioning the efficiency.
Store -> Samples -> Home -> Store -> Samples -> Home… infinitely.
Lowe’s, along with Pinterest and Microsoft, is building the opportunity to see what the remodel would actually look like, from the comfort of a VR exhibit in the store.
The future is here, friends. You don’t know what is going to come next so the only way to keep up is to plug-in and prepare yourself. The relationship between change and time is exponential (tweet this!). Attending conferences like Gigaom AI is a refreshing peek into the future, and provides incredible networking opportunities for players in the space. Get out there and make some waves, people.
Panel: How to Build Better Customer Experiences with AI
- Christian Petersen – Director of Product Management, Comcast Silicon Valley Innovation Center – and a Dane! 🇩🇰
- Mike Kail – Cybric, Cofounder and Chief Innovation Officer
- Rob May – Talla and Publisher of Technically Sentient, AI Investor
- Mark Thiele – moderator, Chief Strategy Officer, Apcera
The session started with a discussion of Microsoft’s chatbot, Tay (@TayandYou), turning racist within 24 hours. Not the greatest customer experience, but an interesting experiment nonetheless. Then there’s Alexa and Google Home ordering things because of prompts they hear on TV. For most people today, a controlled environment is likely the best use for AI, providing the best overall experience.
Problems with AI:
Mike – AI has certain culturally challenges. People don’t think about the small business aspects on a first rollout, and then building up from there. People typically boil the ocean with resources trying to be a revolutionary first to market.
Christian – Comcast has over 10M voice remotes in the market. A takeaway – it’s really tough to reach people how to use new items and computers learn much more easily. Let it be a conversational process, and the adoption / use cases will increase significantly.
Mike – Payment issues and security concerns will arise out of AI, connected homes, and IoT build outs.
Christian – Communication needs to be coordinated across different channels. Over text, social, email, and so forth. How do you make sure that all bots know what the other bots know?
Rob – Food for thought… the roll out will be fast. People are willing to trade in privacy for efficiency, but there will be backlash. Will bad guys be able to create a bot to make you disclose information? Yes, something like this will probably happen. Rapid adoptions inherently create mass confusion and backlash.
Opportunities for AI:
Mark – There’s an opportunity to remove impatience with AI — Call center hangups due to miscommunications and reroutes.
Christian – low hanging fruit for AI – all functionality behind-the-scenes. Finding operations and processes that might go wrong before they have the opportunity to go wrong. “Behind-the-scenes” because it’s not a part of your work, but layered-in and will surface when necessary.
Mark – How many people look at Yelp to make decisions? Choose food based off Stars that other people voted on. There’s a great AI opportunity for Yelp to build profiles based on how you look for food, recommend products and services based on how YOU view that product/service versus what others think.
Mike – Opportunities for Google and Amazon to let AI speak more to (and with) kids, who LOVE to have fun and play with the products. Huge learning opportunity.
Mike – Let trained AI bots automate security threats (speaking to payment security issues). Similar to a Microsoft Word’s Spell Check.
Mike – Focus on user-experience as opposed to enhancing AI just for the backend sake.
Future AI related topics:
- Rob – conversational interfaces. Power users don’t like to user natural language. Full language interfaces are good for remembering syntax. Early adopters complain about typing full sentences, vs. “HK” as a code word or shortcut. Interfaces may be more keyword and acronym-driven.
- Mike – We really need Urban Dictionary understanding.
- Christian – Depends what we consider “natural language.” Look at the way young people chat today. Another idea – the combination of voice and chat AI. Think about pressing the same combination of 5 buttons at 11pm every night, this eventually turns into a 1 button prompt when the AI learns.
A favorite quote of mine from Christian –
The day people start calling Alexa “she” is when we’re on the right path. [tweet]
When people want to have conversations and share aspects of their day with these bots. People need to have relationships with bots, meaning that the bots are required to be intelligent.
Audience question: How far from one bot that rules them all? Speaks to all other bots?
Christian: I’m not comfortable (yet) with Alexa knowing Amazon balances and complete order info. But soon we will talk to one bot all day for everything.
This immediately made me think of the Black Mirror episode “White Christmas” where Jon Hamm transferred peoples’ brains into tiny robotic replicas of themselves which they could control as personal assistant-esque gadgets, but Jon’s character could control their life span. I highly recommend binging the series.
On young people’s fluidity with bots….
Christian: Today you’re seeing big companies doing a land grab right now. Same happened with Mobile. Get locked into a company like Verizon’s small garden, and stick it out with them. Tough to say what’s going to happen in the space.
Mark: I see a likelihood Facebook creates a support center AI soon, nothing much deeper than this yet.
In conclusion, it’s in your best interest to start by using AI for simple cases to solve backend inefficiencies. From there, slowly roll out to consumers in more public-facing activities and engagements, slowly building trust and comfort. Take care of the core experience first, and manage your customers expectations. Rule of thumb: whenever your AI bot doesn’t know an answer, be sure it smoothly transitions to the right person. Fascinating panel session reminding us all to continue focusing on AI use cases most beneficial to consumers and providing meaningful, impactful AI experiences for all consumers.
Over and out! Thanks Gigaom!
[cover image credit to Gigaom, logos credited to Legal Robot, LeadCrunch, Cortex, Netra, and Kylie.ai, Black Mirror screenshot credit to IMDB, other images taken by Steffan Pedersen on an iPhone 7+]