Ekram Alam, CEO and Co-founder of MindPortal – Interview Series

Ekram Alan is the CEO & Co-Founder of MindPortal, a company that is building the future of human-AI telepathy, by developing non-invasive neural interfaces that transform how humans interact with AI. Their mission is to streamline communication by allowing users to interact with AI through thought alone, bypassing conventional methods like typing or speaking.

One of your proudest moments is when you dropped out of medical school, could you discuss this fork in the road, and when you realized entrepreneurship was your path?

Dropping out of medical school marked a significant turning point in my life, not because I was headed towards entrepreneurship in the traditional sense, but because it represented a profound shift towards living in alignment with my authentic self. This decision stemmed from a realization where I saw that my previous path was not one of personal choice but one shaped by external pressures and expectations, primarily to please my parents.

My journey into what some might call ‘entrepreneurship' is better understood as a quest for authenticity and impact. Influenced by visionaries like Michael Jackson, Steve Jobs, and Walt Disney, I recognized my deep resonance with creativity, innovation, and making a significant impact. This realization wasn't about adopting a new label but about aligning my actions with a vision of advancing humanity toward an ever-expanding future.

In exploring this path, I became interested in the potential impact of ideas like psychedelics, which, coupled with readings like “The Singularity is Near,” helped me see the intersection of my personal passions with the broader trajectory of technological and human advancement. This wasn't about fitting into the existing mold of a company or using capital in the conventional sense; these were just tools at my disposal. My real focus was on launching projects that serve the advancement of humanity, leveraging existing current paradigm structures such as a company structure and capital as off-the-shelf vehicles towards this vision while staying open to the emergence of new paradigms that might better serve our collective journey.

The essence of my endeavor is thus not entrepreneurship per se but a commitment to contributing to the infinite game of human progress. This game involves expanding the frontiers of our capabilities, experiences, and understanding in a way that transcends current societal and economic structures. It's about playing an active role in the evolution of humanity and technology, with an eye towards a future where we continually redefine the boundaries of possibility.

To expand on this, my departure from medical school marked not just a divergence from a predefined path but an initiation into a profound exploration of existential alignment and purpose. This journey was less about rejecting a traditional career and more about awakening to a holistic vision of progress and potentiality that transcends conventional labels like ‘entrepreneurship.'

In essence, my voyage is one of aligning with the macrocosmic evolution of complexity itself—a narrative where the universe, through an inexorable process of compounding complexity from simplicity, has birthed consciousness and, subsequently, technology. This trajectory suggests that technology, as an extension of natural evolution, is set to catalyze the next leap in the universe's expansive odyssey.

Through this lens, the societal constructs of ‘company' and ‘capital' are merely provisional tools—artifacts of our current epoch utilized toward the ultimate end of advancing human and technological potential. This perspective posits these tools not as ends in themselves but as malleable means, adaptable to the evolving needs of humanity's quest towards infinite frontiers.

The core of my mission, therefore, is to contribute to this grand tapestry of universal evolution. It is an endeavor that views human progress and technological advancement not merely as goals but as manifestations of a deeper, cosmic game of exploration and expansion. This game is governed by the principles of perpetual growth, transcendence, and the unending quest to explore and actualize the boundless possibilities that lie before us.

This mission is rooted in a conviction that humanity's journey is intrinsically linked to the broader evolutionary narrative of the cosmos. It is a commitment to actively participate in steering the collective trajectory towards a realm of infinite possibilities, where each discovery and innovation serves as a stepping stone to further expansion and understanding.

In embracing this path, I am driven by a vision that is at once deeply personal and universally resonant—a vision that seeks not just to navigate but to contribute to the unfolding story of the universe. It is a path that demands a reimagining of our roles and responsibilities within this grand narrative, inviting us to view our contributions through the lens of cosmic significance and boundless potential.

Could you expand on how reading Ray Kurzweil’s book “The Singularity Is Near”, led you to begin considering the far future and unlocking some new ideas?

It’s a book that explores exponential technological advancement and its implications for humanity. The Singularity term itself has connotations to an event horizon, a future whereby predicting what’s possible becomes challenging because of how profoundly different things will be. Even if we take a moment to imagine our wildest dreams and imaginations of what the future might be like, the actual future will likely be far more beautiful, wild, intense, and an experience we simply cannot fathom with our current primate biological brains. To me, it represents one of the greatest books ever written and I enjoy envisioning the future of what frontiers of human advancement I’m excited for and how technology can enable serving the proliferation of this advancement. The bonus of advancement being exponential is that in terms of absolute time things will happen far faster than we realize or even can intuitively compute in our brains. It’s also an infinite positive sum game of ever-expanding human potential, capabilities, and experiences. In a sense, it’s a book that resonates with me because it’s pro-human, it’s about human advancement, and technology is an enabler of that advancement – technology is not the goal, it’s a means to an infinite expansion of humanity itself. From living forever, to mitigating suffering, to expanding our minds and intelligence, to exploring the universe to infinite more unimaginable expanses of this game and delight that is existence – this book resonated because this is something I love and live for and am excited about and want to contribute to human advancement as part of our collective journey in this tapestry of reality and existence.

In a future where AI will inevitably surpass humans, why is it important for MindPortal  to build technologies to enable humans to more intimately communicate and interact with AI?

Human biological evolution, including our brains, has plateaued. AI, on the other hand, is exponentially increasing in intelligence and capability. It will surpass humans. And this is great, but I want us to have the option of going along for the ride (if we want to – my perspective is that individual humans should choose whatever they want to do). But what I mean by this is a human-AI merger – we become “one” with AI, and therefore, as AI becomes infinitely more capable, so do we because it’s not two distinct entities – our brains and minds are, in essence, increasingly nonbiological.

So, in essence, merging human minds with AI is important. It’s important as it enables us as humans to keep pace and continue to be at the frontier of this cosmic game of expansion, innovation, and advancement, and it’s also beneficial in that it helps to mitigate risks that come with runaway AI. So it’s both a potential for upside and minimisation of risk. It’s a boon as we connect and expand what humanity even is to new frontiers. The definition of humanity, I imagine, will be expanded to include biological humans, post-biological hybrid humans, and pure AIs. All of this is humanity.

MindPortal launched with my vision in mind of a human-AI merger, and it exists to at least contribute towards the foundational aspects of this merger. A merger is a spectrum – two separate entities increasingly connected when brought close enough together to become a unified entity. The cells in your body communicate, but cohesively they form organs and organ systems. By enabling better human-AI telepathic communication, which is the current focus at MindPortal, it sets the foundation for a better communication relationship between humans and AI, even if at present we generally think of humans and AI as distinct but it also paves the way for a deeper understanding of how the human brain communicates, and how AI can be built in a way that operates in tandem with the brain. This is groundwork for the future where these insights can contribute to human-AI merging where the communication between the human mind and AI will be so seamless and high bandwidth that it represents a cohesive internal unified communication with self aka there is no human communicating with AI, but rather a unified entity communicating with its own expanded mind.

So MindPortal exists to enable better human AI telepathy because in the short term high bandwidth comms is good for spatial computing and LLM interaction – it’s better than a mouse or keyboard and represents an expansion of the bottleneck in terms of communicating with powerful tools.

But long term, it represents groundwork and sets the stage for human-AI merging by enabling a deeper understanding of human-AI symbiotic relationships.

Can you discuss the technology that is being developed, and how it differs from some of your biggest competitors such as Neuralink?

Our technology distinguishes itself in two fundamental ways. Firstly, unlike Neuralink's invasive approach, we utilize non-invasive methods. Our primary focus lies in telepathy, enabling users to communicate with AI through imagined language directly. In contrast, Neuralink employs invasive electrodes embedded in users' brains, utilizing a surgical robot for quicker and more precise electrode implantation to enhance spatial resolution.

While invasive electrodes have been demonstrated in academic labs before, Neuralink's innovation lies in the development of a surgical robot to expedite the implantation process. This approach allows them to position electrodes closer to neurons, enabling more direct recording of neuronal activity. Their emphasis on spatial resolution is primarily geared towards medical applications, for example in patients with conditions like Parkinson's.

In contrast, our non-invasive sensor technology is designed with a consumer-oriented focus, catering to a broader audience beyond medical patients. Additionally, our understanding of the human brain recognizes the distinct roles of the neocortex layers. Neuralink's invasive electrodes primarily reach the deeper layers responsible for functions like dopamine production, often relevant in conditions such as Parkinson's.

We acknowledge the distributed nature of language and semantic information in the neocortex. Invasive methods would require multiple surgical electrodes and the requirement for surgical holes drilled into the skull to effectively access this information. Our non-invasive sensors cover larger areas of the brain, allowing for more effective recording of functional dynamics crucial for imagined speech and other areas of interest. This approach eliminates the need for invasive procedures and enhances the overall usability and accessibility of our technology for a broader user base.

What are some of the biggest challenges behind building a brain-computer interface (BCI)?

Our research and development has encountered various challenges and uncertainties, showcasing the novelty of our approach. Here are some key hurdles we have addressed:

  • Experimental Paradigm Design: We faced the challenge of determining the most effective experimental design for eliciting detectable neural signals related to semantics and language based meaning using non-invasive equipment. Drawing insights from neurolinguistics and leveraging our team and collaborators expertise, we designed an experimental paradigm to optimize the chances of identifying discoverable neural patterns associated with imagined speech.
  • Repeatability of Results: Imagining speech produces low-intensity signals in the brain, making it challenging to distinguish from the inherent noise. Additionally, human brain states vary throughout the day and across different days, raising concerns about the repeatability of signals across different recording sessions. This uncertainty persisted in both our word decoding and sentence decoding experiments.
  • Signal Processing Pipeline for Optical Neuroimaging Data: Non-invasive instrumentation like functional Near-Infrared Spectroscopy (fNIRS) introduced various sources of noise in the signal, particularly when attempting to identify different sentences with distinct meanings. The decoding of sentences had not been demonstrated previously with fNIRS technology. Moreover, demonstrating semantic representations had only been achieved with higher resolution fMRI, which is non-portable, rather than with the portable technology we employed.
  • Machine Learning Model Development: Developing a machine learning model capable of classifying different sentences in real-time posed a significant challenge. The task required extensive data science experimentation to identify the optimal model architecture capable of effectively differentiating between neural patterns corresponding to different imagined sentences.

Addressing these challenges has been integral to the advancement of our R&D, pushing the boundaries of what is achievable in the field of non-invasive neural interface technology.

Everyone’s brain is different, how is machine learning used to identify thoughts?

MindPortal acknowledges the unique nature of each individual's brain while recognizing fundamental similarities in speech production and processing networks. In semantics, the neocortex serves as a common ground for representing meaning.

Through our neuroimaging technology, we capture images of brain activity when users imagine different sentences, revealing distinct patterns of activation. By collecting data from multiple participants, we are training a machine learning model to create a generalized model. This approach is akin to image analysis in computer vision, where a model is trained on diverse images to identify patterns, such as recognizing pictures of a bear despite variations in context, shape, or color.

This generalized model can then be fine-tuned through a calibration experiment, accommodating subtle differences in a new user's brain. This process allows the model to adapt, ensuring effective recognition of imagined sentences. In essence, our methodology mirrors the principles of image analysis in computer vision, demonstrating the versatility and adaptability of our neuroimaging technology in understanding and interpreting the neural patterns associated with imagined speech.

Can you discuss one of your guiding principles which is: “what is the Experience We Want a Human to Have”. What specifically is the experience you are aiming for?

In the short term, we're aiming for human-AI telepathy at MindPortal. A human thinks, and something happens. The experience I want is magical. Godlike. You think, and it becomes. You want something done in your environment, and it happens.

The experience should be intuitive and therefore require minimal training to use. You just naturally imagine speech in your mind e.g. you might imagine a sentence or inner monologue in your mind such as “Show me the best restaurants in my area”, and it instantly happens. The experience will be private, hands-free, accurate, intuitive, and in real-time.

The wearable we're designing at MindPortal will be lightweight, fashionable, stylish, comfortable, and integrated into your daily life. In the future, this technology will integrate with AR hardware like headsets or glasses. Imagine yourself projecting a hologram of a rocket, and you think to yourself, “What changes do I want to make to this rocket?,” and the design will auto-shift according to your thoughts and desires while you lay back on your sofa.

MindPortal is currently working on mentally interfacing with Large Language Models, can you discuss your progress in this?

MindPortal has made significant progress with demonstrating human-AI telepathic communication between a human user and Large Language Models such as chatGPT. We started by successfully classifying sentences and words using non-invasive technology, specifically through optical neuroimaging. This breakthrough allowed us to decode sentences with different meanings directly from the human brain. In early 2023, we demonstrated our ability to decode individual words like ‘yes' and ‘no' using a specialized headset, enabling users to interact with screens by imagining ‘yes' or ‘no'—the first-ever demonstration of human telepathic interaction with ChatGPT. Building on this success, we advanced to decoding full sentences from the human brain. Users could imagine sentences from a set of pre-trained options, and our machine learning model recorded their real-time brain state, predicting the imagined sentence, which was then communicated to ChatGPT for a response. Currently, we are developing our next milestone, set to be unveiled in Q3 2024. This advancement will showcase a significantly larger vocabulary of sentences, covering the range of expressions users convey in everyday speech.

What is your vision for the future of BCI?

In the short term, our vision for brain-computer interfacing focuses on two areas. One area is spatial computing integrated input. Imagine an immersive experience in virtual and augmented reality, where executing commands is as simple as thinking. Your thoughts directly shape the virtual environment, making interactions effortless. The second area extends to the telepathic integration with Large Language Models (LLMs) and AI assistants like Alexa, Siri, or ChatGPT. The ability to receive responses and execute commands through thought alone represents a significant shift in human-AI interaction, offering unparalleled convenience. The two areas are likely to increasingly converge in the future, whereby you can be in an AR experience and navigate your spatial computing ecosystem with the help of an AI assistant with whom you telepathically communicate.

Looking further ahead, our long-term vision involves a profound human-AI merger. We foresee a continuum where minds seamlessly integrate with AI, where humans increasingly move from being predominantly biological to predominantly non-biological. This evolution positions us as post-biological humans, with individuals having the autonomy to determine their existence. Crucially, this transformative shift is voluntary, respecting the preferences of those who may choose not to embrace this hybridization with AI. Our goal is to empower individuals with the freedom to decide the extent of their integration with AI, fostering a future where human-AI collaboration is both versatile and customizable.

Thank you for the great interview, I look forward to following your progress, readers who wish to learn more should visit MindPortal.

The post Ekram Alam, CEO and Co-founder of MindPortal – Interview Series appeared first on Unite.AI.

文 » A