Focused on the Future of AI
Artificial Intelligence will transform the world in ways and to an extent that no technology has ever done before. No time in history has ever had such a perfect confluence of technologies, needs and opportunities–all converging upon and underpinning AI.
The perfect convergence of timing and synergistic factors include:
- Artificial Intelligence – Out-performing prior human or computer approaches
- Data Resources – Providing base to power Deep Neural Networks training
- Computing performance – GPU’s and ASIC’s providing processing for AI
- Blockchain Technologies –Enabling highly secure distributed applications
- Quantum Computing –Convergent in capabilities to geometrically
- Global Usage Explosion –Projected doubling of online users in 8 years
The result of this is what is being called “the next era of human evolution.” When this digital evolution is coupled with what could be a several hundred-fold increase in quantum computing performance growth per year, calling the coming period a revolution in human-enabled technologies may be an understatement.
The unprecedented number of industries and knowledge domains that AI addresses requires more specialized, and at the same time, more interdisciplinary knowledge than ever before. As such, the more people actively involved in guiding this revolution, the more likely it will represent a broader base of needs. Since we have no prior historical precedent to enable us to understand the full nature, capabilities or implications of this confluence, we need to understand and maintain involvement to help ensure it results in what we believe is positive change.
OUR GOALS & OBJECTIVES
Growth – Jobs, education, economic
Knowledge – Theoretical, practical, applied, human factors
Problem Solving – AI Development & consulting services
Improvement – Community, health, security, human condition
Seeflection, together with its constellation of alliances and partnerships, is leveraging these capabilities to create human needs-driven solutions. We are also focusing on the challenges brought by these powerful new capabilities and their resulting changes, which will impact virtually every area of human endeavor and our civilization.
Primary Areas of Focus
Primary / Applied R&D – Deep Learning / Machine Learning & Quantum AI
Core & Enabling Technology Development – Advanced AI Media Lab and vertical R&D
Strategic Research & Development – internally and with key partners
The launchpad for a productive future
Seeflection is focused on helping to ensure that Artificial Intelligence, Quantum-AI and Blockchain-AI are leveraged for our collective benefit. To that end, we work to democratize access to, and information about, AI and how it can be applied in safer and more beneficial ways. We believe that this historically unique confluence of mutually-supportive technologies provides a powerful launchpad to improve many aspects of our lives and businesses. We believe that this may well be a once in civilization chance to help create the basis for a completely new type of business, lifestyle and social fabric, which should not be ignored or squandered.
We have entered the first decade of multiple global doublings
As a result of AI and the prior generations of digital growth, the exponential growth of digital technology is about to sweep most brick and mortar businesses and approaches along with it.
Several projections expect a doubling of the world’s gross domestic product by this coming decade. AI and associated robotics, workflow optimization and process automation are similarly projected to increase margins by upwards of several hundred percent. Quantum computing and Blockchain will affect everything from science and health to commerce and financial services and many other areas to a comparable or greater degree than AI alone. Thousands of new types of products and services are being developed already, many of which create entirely new markets and jobs/business opportunities. However, without proper planning based on a collective, collaborative effort among those responsible for this transition, we could miss much of this amazing opportunity to address a wide range of societal, economic and other issues.
Maximize opportunities, minimize potential downsides
Something that affects virtually every area of our personal, business and other activities requires planning to leverage the opportunities and minimize the potential negative aspects. Given the scale and breadth of this change, the ramifications of not planning realistically for this and leveraging the huge opportunities that are being presented would be disruptive and wasteful at best and catastrophic at worst.
The necessity and benefits of joint collaboration and planning
We believe that a critical part of our mission at Seeflection is to help to focus efforts to productively our new resources and opportunities in. Given all of the unique ways that AI will impact each industry, a positive transition requires: interdisciplinary planning, knowledge and skills dissemination and development, and above all — an intention to focus our collective resources to help ourselves.
Seeflection’s Action Plan:
Identify and support strategic AI-enabled opportunities to create directed growth
Leverage our knowledge, labs, studios and alliances to develop training and assist in developing education curriculums and initiatives.
Work together with local companies, government, educational and training centers, to support initiatives that bring projects, jobs, capital and related opportunities to the region.
Expand local research and development alliances with key teams and companies in data science, AI, quantum computing, blockchain and strategic vertical markets.
Develop/continue to evolve relationships with Open AI Initiative, Google, NVIDIA, IEEE, MIT and others to bring AI ethics, scenario modeling, planning and resources to the region and AI markets.
Help develop local and regional initiatives to support efforts for our population’s transition
Create local AI and related technology incubators, R&D centers, educational and vocational training/retraining centers to help enable Arizona leverage massive jobs creation, growth and related opportunities.
Create AI-assisted online and associated information resources to help guide, support and train residents take positive advantage of this AI-enabled revolution, through opportunities in business, cost savings and evolving and enhancing skill sets.
Help create “Centers of Excellence in Innovation” in key strategic fields, acting as resource-multipliers for our human and capital resources, while providing a model to expand alliances and opportunities beyond the region.
Work with utility and resource providers to identify and leverage AI-enabled Blockchain and related AI systems to increase corporate margins, while reducing consumer costs and risk factors.
Some current projects
Rapid AI development environment for Deep Learning networks
- fast, efficient, works with standard technologies
A growing set of API’s and libraries that work with modular Tensorflow and Keras scaffolding to support static and dynamic computational graphs for deep learning development. Created to save time and standardize and optimize them for Tensor RT3 accelerated training and inference.
Blockchain secured Point-to-point/multi-point audio/video multicast architecture
- Ultra low latency, advanced media delivery
Leverages RTMFP and Kovri over a connectionless protocol to support blockchain smart contract extensibility that enables pre-defined media/data lifespan, compensation guarantees and data/media asset protection, identity/ownership validation and creator-defined transport, usage and delivery contracts.
Hyper-accelerated GPU-based Blockchain hashing system
- enables what previously required racks of ASIC/GPU hardware
Simple set of libraries optimized for GPU CUDA/TRT3 execution to (initially) provide 1,240 T-Hash/Second per quad DGX
Kovri and I2Pd AI extensions, leveraging deep learning (GAN/ Auto-encoder ensemble AI)
- to enhance privacy and protection of distributed data
Enables strong privacy via distributed redundancy using Blockchain, with the I2P encapsulation with AI-driven smart contracts extensibility
Dynamic time-scaled tensor propagation for deep learning networks
- alternative to back propagation with efficient solutions matrix
Uses fibonacci period embedding execution latching to increase partial differentials convergence rates, with intrinsic local and global min/max identification.
CAPSNet TensorRT3 optimized ensemble AI network architecture
- supports high performance deep learning
Uses a one-shot learning together with optimized ensemble inference engine to enable realtime learning and inference operation.
AI-enabled synesthetic mapping and retraining multimedia bot
- work in progress being designed to bridge disparate sensory areas to support neurological development
Potential applications include physiologically, mental/emotionally challenged and injured individuals to aid in recovery and independence.
Projects & Collaboration
Seeflection is focused on developing synergies–both internally and externally. In terms of technologies, the timing of AI, Quantum Computing and Blockchain technologies seems almost scripted. I don’t think I am aware of a time where there has been such a parity between our need and technologies. The mutual support between the areas of technologies, and the way in which they inherently address the gaps of the others makes it almost seem like it was written to happen this way.
Similarly, the Seeflection team has been focused on interdisciplinary areas that complement and support these technology areas, and have been developing AI and related software systems for almost 30 years.
Ongoing projects
Convolutional neuron (Layer 0-1) aggregation, using edge-biased activation profiling
- CNN mod for increased low-contrast/low lumen edge/object extraction
Based on neuron activation profile modeling to increase image recovery at low lumen levels using AI drive collapse of chroma in order to integrate with luma tensors to improve object identification rates in low contrast/low luma environments.
NodeJS driven Hadoop/MapR Blockchain API and IO libraries
- to combine I/O and data scaling with blockchain security and usage programmability benefits
Adds new types of AI and traditional data management and processing capabilities for large (PB+) datasets.
Sparse matrix correlation inference library
- helps find hidden correlation between loosely coupled domains
Initially conceive to explore correlative relationships close to Nyquist limits, but may possibly be expanded to help with general optimization convergence.
RTensor3 Blockchain-SQL extensions that enable secure GPU accelerated data handling and management
- GPU accelerated blockchain-SQL extensions
Accelerated set of MongoDB NoSQL extension being integrated to enable high performance with large sets of semi-structured, unstructured data using HADOOP-Blockchain-data structures.
Euler-spiral / differentiable spiral inference engine
- to help support autonomous driving and racing performance and general safety optimization
Uses first, second and third PDE of G-forces to help optimize path a vehicle follows, based on GAN AI topology that compares Euler-spiral with first order parabolic 1-O/2-O differential equations.
Phi-rotated tensors library extensions for quantum-classical AI operation.
- proposed probability/phasing quantum-AI integration architecture
An approach based on a complex phased-probability tensor integrated via directed acyclic graph via resulting dot products.
Indirection based deep learning tensors, using RT3 GPU optimization
- for data handling flexibility with increased performance
High performance applications with decreased training pool requirements, and support for inductive inference.
Bridge-weighting deep learning architecture
- provides bias profiling support for hyper parameter optimization
Based on neural activity pattern studies. Initial applications being explored include rapid hyper-parameter optimization
– Meaningful Quotes –
“With artificial intelligence, we are summoning the demon.”
“Our technology, our machines, are part of our humanity. We created them to extend ourselves, and that is what is unique about human beings.”
“By far the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.”
“You asked the impossible of a machine and the machine complied.”
“A year spent in artificial intelligence is enough to make one believe in God.”
“I did not like experimental psychology. The kinds of theories they were willing to entertain were hopelessly simple. So, I became a carpenter for a year. I wasn’t very good at that, so I did a Ph.D in AI.”
“As soon as it works, no one calls it AI anymore.”
Discover The Future
We are a multi-disciplinary laboratory committed to focussed and efficient drug discovery
