AI-powered technologies, solutions, and business models are now permeating almost every known industry from healthcare and pharmaceuticals to finance, security, and military defense; exponential technologies, especially those powered by AI, present a variety of new and exciting opportunities/products individuals and enterprises can employ for their benefit.
However, few people truly understand the breadth of possible applications related to AI-driven technologies and may struggle to find companies with original products or services without pre-conceived industry knowledge.
Throughout this article, I will present a selection of companies I believe are pioneering the most innovative, novel, and useful AI-powered products (this by no means encompasses everything that exists within this industry). Some of these companies may already be on your radar, while others will likely come as a welcomed discovery.
NVIDIA, founded in 1993, is now one of the most globally well-known AI software and hardware manufacturing companies. They specialize in the engineering of graphics processing units (GPUs), system on-a-chip units (SoCs), and application programming interfaces (APIs). However, over the last few decades, NVIDIA has distinguished itself from other companies in this domain.
In 1999 NVIDIA launched the GeForce card, a GPU that rose above all other competition with its impressive abilities for 3-dimensional graphics generation, and lighting dynamics. GeForce NOW is one of their latest developments in this product line – any users that download the software gain access to a vast cloud-based gaming inventory – NVIDIA also offers different subscription packages gamers can upgrade to for longer session lengths, better resolution, and higher processing speed.
In 2006, the company released its own coding language, CUDA, which programmers could employ to customize their GPUs and run parallel processing algorithms – the ability to run these algorithms greatly increased processing power, and allowed programmers to develop sophisticated techniques for data visualization, simulation, and modeling. CUDA is now a well-known coding language; it is taught throughout several universities worldwide and is comparable to C++ and Python.
In fact, the company is currently using this technology to build EARTH-2, an AI-driven supercomputer that aims to predict and curb the effects of climate change at scale. This is but one example of the company’s versatility; they have applied their technology to numerous other industries as well:
- Sustainable development
- Military defense
- Urban design and infrastructure
NVIDIA is currently partnered with Amazon Web Services (AWS) and Google – these two major players in the tech scene represent only the tip of the iceberg, and it is likely that as the company facilitates its expansion into more industries, prominent partnerships will continue to blossom.
DataRoot Labs is a data science and AI consulting company that was founded in 2016 by Max Frolov and Yuliya Sychikova. They offer services throughout the three primary domains: 1) AI solutions development, 2) provision of AI and ML engineering teams and 3) startup consultation and venture services.
In AI solutions and development, DatatRoot Labs offers their customers the chance to gain a free initial consultation – from there, the company helps programmers evaluate their model design and training procedures, solution architecture, model development, and deployment as well as any other relevant characteristics, in addition to the provision of a full software stack. Most recently, they have helped build an empathetic virtual assistant that aims to reduce loneliness among elderly citizens using advanced NLP technology.
DataRoot Labs provides an initial core AI and ML engineering team that helps companies streamline their hiring process by building data science teams of their own. DataRoot Lab’s teams also provide ongoing tech consultation, HR, legal, and financial support, as well as tips for better office management and company culture. One of their recent projects in this domain involved building computer vision algorithms that aided autonomous drones in their abilities to identify and detect faults or malfunctions in solar panels.
Finally, DataRoot Lab’s startup consultation and venture services offer potential clients an initial free consultation that aims to consolidate relevant tasks within a respective business model. From there, the company helps clients with investor outreach, deal negotiation, financial model generation, and any other hurdles the start-up may encounter. Recently, DataRoot Labs worked with the hiring platform dhired.com to develop a candidate evaluation model that scored professionals based on their learning and technical skills.
AssemblyAI specializes in the provision of speech-to-text APIs for enterprise use; ultimately, they aim to provide developers with advanced deep learning technology that remains accessible and easy to use over time. With over 63 million dollars in funding, much of it coming from well-respected investment firms such as Accel and Insight Partners, AssemblyAI is not a company to shy away from, especially when considering the potential utility of their products.
Currently, they offer two primary products in the domains of core transcription and audio intelligence.
The main idea behind core transcription is to provide developers with the tools they require to generate text with human-level accuracy from live audio, as well as pre-recorded audio and video. Importantly, the API AssemblyAI has built is extremely robust; it is scalable, secure, and compliant with GDPR regulations. Some of its main features include:
- Real-time transcription: the ability to convert live audio streams into text within milliseconds
- International language support
- Confidence scores: the attribution of translation/transcription accuracy scores to individual words in text post-processing
- Async transcription: the ability to convert audio and visual material into text within a matter of seconds with a distinctly high degree of linguistic accuracy
- Custom vocabulary: tailoring vocabulary to business/marketing needs
For those who are interested in conducting higher-level linguistic analysis, AssemblyAI also offers API audio intelligence tools, upon request, that evaluate the sentiment of a given phrase within a body of text as well as entity identification, or the ability to pinpoint names, locations, concepts or objects.
AssemblyAI’s audio intelligence imbues developers with the ability to construct applications with impressive features that allow for a deeper understanding of synthesized text content. Beyond sentiment analysis and entity identification, they also offer tools for
- Content moderation: the ability to identify potentially sensitive or harmful content
- PII Redaction: the ability to selectively pinpoint and subsequently remove any information that can identify individuals (e.g. SSN, IP address, bank account number, etc.)
- Topic Detection: the ability to identify topics in video/audio and then generate topic labels using IAB Taxonomy
AssemblyAI is an exciting company that is worth keeping an eye on, especially since they are currently pursuing new initiatives in audio intelligence that include detailed emotion detection and analysis, intent recognition, translation, and ad detection.
Kore.ai aims to deliver and develop the next generation of AI-powered virtual assistants for enterprise-level utitily. Using state of the art Natural Language Understanding (NLU) technology, Kore.ai helps their customers streamline engagement with relevant stakeholders and transform the way in which human-machine interactions are automated.
Kore.ai offers a variety of products that span across 5 different domains:
- Customer Experience
- Employee Experience
- Agent Experience
- Search Experience
In terms of customer experience, they offer thee AI-powered virtual assistants: SmartAssist, BankAssist and HealthAssist. Supported by the Kore.ai cloud platform, XO, SmartAssist acts as a holistic AI-native contact center that facilitates customer and agent engagement by identifying consumer intent and rerouting relevant queries to appropriate branches of management. Alternatively, BankAssist provides customers with a multi-channel interactive conversational AI platform that allows them to check their balance, conduct P2P transfers as well as regular transfers, schedule bill payments, view transaction and spending history, in addition to several other finance-related functions. HealthAssist functions similarly, in that it streamlines voice and text communications with patients, providers, and caregivers across every enterprise domain.
Kore.ai’s employee experience products are IT Assist, HR Assist and WorkAssist. IT Assist improves employee’s abilities to request and resolve IT-related issues throughout enterprise domains by building omnichannel support systems, account access management tools, automating ticket organization and flow, as well as interactive communication in response to knowledge searches, among several other functions. HR Assist provides employees with many of same tools as IT Assist except within the HR branch of a business. WorkAssist, on the other hand, is a tool that enterprises can use to reduce the frequency and prevalence of data silos, centralize their data to streamline the collaborative virtual workspace, and ensure that business teams, tools, and information are integrated in context-relevant ways.
Kore.ai’s search experience tool, SearchAssist, helps users conduct intelligent searches across their respective web and mobile channels. It provides a personalized experience whereby customers receive individualized product recommendations and search query results which are updated over time to reflects changes in preference using ML capabilities. Conversely, Kore.ai’s agent experience tool, AgentAssist functions as a personal virtual assistant that monitors and tracks agent-customer conversations and interactions, while also conducting sentiment and intent analytics.
To wrap it all up, Kore.ai provides their customers with a comprehensive all-in-one infrastructure with interpretable AI models, smart assistant building and data visualization tools, as well as a no-code builder. Kore.ai’s versatility and breadth of products make it a prominent conversational AI company that is worth carefully considering.
OpenAI is a company that currently occupies a place at the forefront of AI innovation and development. OpenAI’s ultimate goal is to build fully functional generalist agents or artificial general intelligence (AGI) for the benefit of humanity. Some readers may already be familiar with their previous language model, GPT-3, which with 175 billion parameters represents one of the largest and most robust language models to date.
OpenAI’s latest product development is DALL-E 2, which functions as a creative image generator. The user can input a written phrase, abstract or concrete, and DALL-E will generate a variety of different images depicting what is written in the phrase. What is especially impressive about DALL-E 2 is its ability to understand the nuances of linguistic structure and semantics, such that it can comprehend context-relevance and produce not only accurate but aesthetically pleasing images.
The advent of DALL-E 2 does not only represent a benchmark in natural language processing and understanding, but also an significant achievement in AI creativity. Conventionally, creativity has been regarded as a uniquely human trait, although DALL-E 2 has begun to shift these boundaries in favor of artificial generalist agents.
FutureAI, much like OpenAI and Deepmind, is attempting to pioneer and develop implementable solutions to AGI. Currently they offer two products: Sallie and Brain Simulator II.
Unlike other available generalist agents (e.g. DALL-E, Gato, GPT-3), Sallie uses mobile sensory pods to learn and interact with a real-world environment. Over time, FutureAI believes that Sallie will be able to cultivate an understanding of complex cause and effect relationships, develop a meaningful comprehension of physical objects, and actively perceive the flow of time by employing its sensory capacity for vision, hearing, touch and speech development. Sallie could be used across a variety of domains, most notably intelligent virtual assisants.
Fortunately, FutureAI also believe in open-source software development and collaboration, which is precisely what they aim towards with their release of Brain Simulator II. This free product is actually a tool-kit that customers can use to build and customize their own generalist agents. The system has a remarkably simple user interface coupled with impressive analytical capacity sustained by a spike neuron engine that can process up to 2.5 billion synapses per second.
Companies that choose to pioneer AGI, if they do so successfully, could create technologies that are as skilled and versatile as humans across virtually every conceivable industry. A successful AGI would likely revolutionize not only AI research, but the world as we know it. FutureAI is a company that anyone interested in the AI landscape should be paying close attention to over the coming years.
Similar to Kore.ai, Boost.ai offers their customers a holistic conversational AI platform that aims to steamline and automate interaction and communication at scale for enterprise utility. Boost.ai primary development imperative is scalability – as an effect, the platform they have designed includes no-code conversation builders and high-level NLP/NLU technology that helps automate virtual assistant interaction across thousands of topics.
Importantly, their conversational AI is both versatile and easily trainable, meaning that Boost.ai’s customers can use integrated analytical tools to gain insights into how they can customize and improve their AI assistant over time. Moreover, the AI assistant is compatible with a wide array of already existing communicative software platforms including Google, Microsoft Teams, Skype, Slack, Salesforce, Facebook Messenger, and several others.
For businesses that are looking to adopt a flexible yet robust conversational AI platform for the purpose of streamlining autonomation, collaboration and communication, Boost.ai is a strong choice.
Planet Labs is an organization that uses AI methodologies in conjunction with satellite imagery data to deliver nuanced insights into the course of change on our planet. The insights they provide can be used to help guide decisions in numerous industries from agriculture, mapping, and government, to military and urban development.
Using a fleet of 200 custom-built and designed satellites, Planet Labs provides clients with high-frequency satellite imaging data that is not available elsewhere commercially. For instance, the company creates remarkably high-resolution basemaps that can track physical change over time in a given location. Importantly, these maps help state and local governments visualize resource allocation and demographic shifts at scale – using Planet’s ML capabilities, governments can then pinpoint trends or patterns that are revealed throughout this data to make more informed sustainable, and humanitarian decisions.
Planet claims to have approximately 1700 images of every location on earth, which it uses in combination with satellite imagery obtained daily, to compose a comprehensive global basemap at the end of each year. This map produces an impressive representation of the entire earth’s surface and allows climate scientists and researchers to visualize environmental change in excruciating detail. Moreover, this can be supplemented with Planet’s spectral maps, which filter out atmospheric effects, providing high-resolution images of natural environments that are ready for spectral and quantitative analysis.
Planet Labs is an organization that is worth keeping an eye on, not just because what they are doing has the potential to revolutionize our existence on this planet, but also because they operate within a very narrow market – in other words, there are still a number of growth opportunities that will likely arise throughout the coming years, especially as the effects of overpopulation, urbanization and climate change become more prevalent.
Neurala is an up-and-coming software development company that specializes in AI-driven automation and computer vision software for quality inspections in manufacturing. Neurala’s VIA software package is compatible with essentially any existing hardware and allows manufacturers to decrease the rate of product defects by improving inspection accuracy as well as the number of inspected products.
The company uniquely provides an explainability service as well. This service allows manufacturers to see precisely what their respective AI ‘sees’ – this results in manufacturers having the ability to tailor and customize their AI in ways that improve and adapt to changes or deficiencies in the production process.
Importantly, Neurala is not just another software development company. The founders, who previously worked in the government research sector for NASA and DARPA, developed and patented the highly successful Lifelong Deep Neural Network (L-DNN) algorithmic design.
Unlike DNNs, L-DNNs do not require the preservation of training data and have much higher training speeds. Conventional DNNs use backpropagation, an algorithm that iteratively adjusts the weights of each node in a given layer on the basis of error classification rates obtained during each successive training phase. While this technique can produce impressive learning capacities, it is extremely sensitive to new data, whereby when it is introduced, all previous information is erased. L-DNNs manage to sidestep this issue by mimicking the computational function of human memory, such that new pieces of information can be introduced in tandem with already existing ones.
Importantly, Neurala’s Brain Builder SaaS platform uses L-DNN technology to help provide potential clients with the ability to derive novel and customized computer vision solutions. Moreover, this software platform integrates data analysis, training, deployment, and tagging functions to appeal to both proficient AI engineers and beginners alike.
Neurala has won numerous awards over the last several years for its products and innovations – the founders and employees also have highly reputable research backgrounds coupled with extensive industry experience and creativity. It would come as no surprise to see this company break into the mainstream very soon.
AIBrain is a daring company that aims to augment human intelligence with the help of AI-powered technology. The ultimate goal of the company is to build a fully autonomous AI whose problem-solving, memory, and learning capabilities equal or surpass those of humans. Currently, they offer two primary products: Tyche and SMILE.
Tyche is designed to be a smart-phone compatible AI robot for children – there are currently five available applications for download on Google, although they are restricted to Android hardware. Importantly, Tyche is interactive; it can display a visual range of emotions in response to certain actions or statements, it can retain both faces and language in memory which enable it to reason, plan and execute/solve novel tasks, and most importantly, it can hold conversations in real-time.
One of the Tyche apps, Build Your Own Intelligence (BYOI), actually allows developers to experiment with the functional capabilities of Tyche by tweaking and customizing models in specific ways.
SMILE, or The Social Mind for Intelligently Learned Emotions, is an AI-powered interactive system that helps streamline social and emotional learning in children. TESS, the virtual agent in the system, possess numerous capabilities: it can listen, converse, plan, recognize, remember and motivate young students, helping them maintain their self-esteem and scholastic performance as they become older.
SMILE, however, is not just another piece of potentially sensationalist technology; it was developed in tandem with neuroscience research that highlights 5 fundamental traits associated with social and emotional learning in children: self-management, self-awareness, social awareness, relationship skills, and responsible decision-making. SMILE, in other words, does not use a set of arbitrary characteristics to dictate its interactive processes – there is a distinct goal in mind that is substantiated by empirical research.
It seems AIBrain has struggled to gain traction in the market over the last several years. However, their human-centric technology and strong mission statement in combination with an expanding and evolving market for AI-driven education indicate that they are worth paying attention to. Personally, I have high hopes for this company, especially since this kind of technology would produce widespread benefits for children with learning or social disabilities.
Sasha is currently pursuing an MSc in Bioethics at King’s College, London. Prior to engaging in his current studies, Sasha was a Division 1 Ski Racer at Bates College, where he graduated with a Bachelor’s in Cognitive Psychology and Classical Philosophy. He is deeply interested in applied ethics, specifically with respect to AI-driven exponential technologies and how they might one day affect humanity