Applied AI Conference by BootstrapLabs
The Applied AI Conference is organized and hosted by BootstrapLabs, a venture capital firm that invests in startups utilizing artificial intelligence in their products.
The conference, held at the Westfield San Francisco Center, addressed interesting questions like “How far are we from Artificial General Intelligence (AGI) and how do we get there?” and “How will AI ‘National Supremacy’ tip the power scale and impact the current geopolitical landscape, as well as U.S. national security policies?”
Sessions covered a variety of areas, including how to:
- Create Humanity 2.0 by fostering AI-human collaboration.
- Ensure that human capital remains the core engine for growth and innovation.
- Feed the nine billion people who are projected to inhabit our planet by 2050 with AI and controlled environment agriculture.
- Protect computer networks, while maintaining privacy with AI.
- Transform the workplace and impact enterprise productivity.
- Develop AI for transportation.
- Turn doctor- and hospital-centric healthcare into a patient-centric service with AI.
- Empower entrepreneurs and accelerate job creation and global economic growth.
Human 2.0 — The next frontier for man
Keynote speaker Dharmesh Syal, partner & CTO of BCG Digital Ventures, talked about the new frontiers of human intelligence and ways to extend human abilities with technology and artificial intelligence.
He claims that, “Within thirty years, we will have the technological means to create super human intelligence. Shortly after, the human era will be ended.”
The next generation of human technologies will disrupt how we live, work and learn: Autonomous systems will operate side by side with humans. In the short term, he expects that robots will be used to work with humans rather than replacing them.
Features of Human 2.0
- Step changes in health, wellness, and social, emotional and behavioral care. Virtualized care with human touch. Neuro, organ system, and genetic tech.
- Life improvements as humans co-exist with autonomous machines that offer degrees of automation and personalization. Connected thinking, feeling, acting.
- Work smarter and safer: Knowledge worker memory and moods. Better sleep and sharper attention. Industry 4.0 with safer human and machine automation.
- Adaptable learning: Paradigm change for learning. Learning behavior feedback loop. Adaptable learning styles, pace and context.
- Behavior data exchanges: Behavior data protected and exchanged. (Blockchain might be a good fit in this field.)
AI And Human Capital
The time to adopt machine learning (ML) and big data has passed. Companies should plan their strategies and retrain their employees as soon as possible. Forty percent of the creative workforce’s time is wasted in repetitive tasks they find demoralizing. These kinds of tasks should be automated by services like chatbots.
Existing backend systems should be made available to new channels and new input types like voice and gestures. These will be important in the near future and will make these systems available to new users.
Also, companies will gain valuable insights from their internal data. If an employee wants to change jobs, companies can predict this six months in advance with their existing data. These kinds of insights will improve corporate efficiency.
AI will become a commodity, as common as electricity. People won’t need to be an expert to use an AI application. Electricity transformed civilization in the 19th century, and now it’s a commodity. Everybody knows how to change a light bulb or how to use an electrical switch — AI will be no different in the future.
On the Road to Artificial General Intelligence
Danny Lange spoke on this topic — his speech was outstanding and inspiring. He explained how he works on breakthrough AI technologies. His approaches include:
- Curriculum learning: Start easy, increment the difficulty and progress on the graduation.
- Memory in learning: It is important that an AI system remembers the results of the prior test; otherwise, it will start over after each iteration.
- Curiosity is a huge factor in Reinforcement learning. If an agent acts randomly, it might take 1000x times longer to train compared to a curious agent, which will try different objects and will have a diverse data set in fewer iterations.
Danny presented many cool demos which he has developed on the unity ai platform. Learn more at unity3d.ai.
AI National Plans, State Funding and Geopolitical Implications
U.S. government agencies showed great interest in the event. Many different agencies participated on the panel where they discussed their work with startups in Silicon Valley. Many agencies have opened offices in San Francisco and the Bay area to work with and invest in startups.
Currently, there are few countries with an AI National Plan, but the number is increasing rapidly as governments recognize the importance of AI for national security and competitiveness.
Europe, China and a few other countries have adopted national plans, with China leading the funding for AI projects. Red colored countries are interested in creating a national plan. Unfortunately the rest of the countries, including those in South America, Africa, Asia Pacific, Eastern Europe and Turkey, do not have national AI plans. The speaker noted that even a country with an AI national plan may not make any significant investments in AI.
How AI is Shaping the future of Manufacturing, Construction & Industrial Design
Manufacturing, construction and industrial design is a sector worth trillions of dollars. Most construction projects are over budget, so improving this sector could deliver huge benefits. Through continuous monitoring, AI can measure when and how long work takes to complete in the field. There are many opportunities to increase efficiency with AI.
In manufacturing, Generative design will be a game changer. For example, using generative design, an automobile manufacturer currently produces some parts that are 40 percent lighter and use 30 percent less material. Generative design factors in materials, manufacturing methods and problems and solutions. With this approach, thousands of prospective designs are generated — only the best designs that address all factors are chosen. All design generation and evaluation is made with the help of the AI.
While AI can substitute humans in certain fields, it is creating more jobs — humans must do the sophisticated work that is is not easily automated. Humans can tackle problems that are not considered viable to solve. For instance, most buildings are similar around the world, but with the help of AI, we will be able to design more specific homes based on specific needs and environments.
In the near future, robots will have a more significant role in production. But training robots in real environments is a slow process. Most of the robot training will be done in simulated environments. AI will make a breakthrough when it uses knowledge generated in one field and applies it to different field.
How AI + Networking can Solve “Unsolvable” Problems
AI can solve problems that are currently considered unsolvable. For instance, you might want to understand what kinds of devices are connected to a network — today there might be thousands of different devices with the number of devices continually increasing, a challenging problem. But because all of the devices use different network interactions, it’s like they all have their own unique fingerprints. With the help of AI, it isn’t too hard to train the system to identify each device from its network trafficking pattern.
Another important aspect of networking is security. But how you can validate the traffic is secure when it is encrypted? You might put a trusted device in the middle and decode, analyze and encode the packet again. But this might cause other security and privacy concerns, and this solution is CPU intensive. With the help of AI, you can extract observable features and train the system to detect unsecured traffic — a more elegant solution.
Enterprise AI Application
The importance of AI has increased dramatically for enterprises, but most companies still lack clean data. Because this is a real challenge, it should be addressed. AI and AI experts should be involved early in the data collection strategy and process.
Because of the way data is currently structured, IT teams waste a lot of time on data wrangling. Companies find it difficult to retain senior data scientists who expect to work on data models but find themselves trying to clean data. Enterprises need to address this problem by teaching junior employees to wrangle data.
Trust will be an important element of AI applications in the near future. They will not be the black box applications of today. Humans need to understand AI in order to trust it, so AI applications will need to be more transparent. A comprehensive system should include human quality control to improve and tune it.
It is a fair statement to say that AI has not yet disrupted healthcare. Trust is more important in healthcare than other sectors — but do doctors and patients want to use AI? Do they trust it? Are they ready? There are no certain answers, but AI is being used for everything.
In healthcare, practitioners and managers want to understand the solution. AI is just a tool and not the solution. The solution must be more heavily emphasized. The healthcare industry is open to automation and new techniques, but the ubiquitous “AI” label discourages trust from both doctors and patients.
We have been hearing about industry 4.0 for a long time now. It should be noted that AI and Industry 4.0 and all its transformations do not have the same effect around the world. Africa is not going to be affected by this transformation in the same way that Europe or Asia will be. Africa’s population is rising faster than that of any other continent. Because Africa doesn’t have established industries, they are going to build their industries from scratch with the 4th industrial revolution. Africa has a lot of potential and presents opportunities for investment.
Management in the Age of AI
Many people believe cloud computing and technological advancements are lowering barriers to entry for the tech business. Anybody can create breakthroughs in the tech industry if they have the right skill set. But AI reduces the chances of success for newcomers. Established companies have a huge advantage because they have the data, a precious resource that has been compared to oil. But the value of the data depends on how you use it and how much you can extract from it.
AI might change macroeconomics as we understand it today. You might remember the classic supply and demand chart from a macroeconomic course. During this session, the presenter observed that the chart is no longer applicable. There is no equilibrium point for a product. All products will be produced and priced for individuals. Most people will pay different prices for the same products. We are getting used to this situation when we use a ride sharing app or buy an airplane ticket.
The Applied AI Conference was very informative and provided good insights into where the industry is going and areas to consider investing in. We can’t wait to share the next AI conference with you.