The AI Revolution: Eight Deadly Trends

Is AI as-a-Service (AIaaS) an emerging commercial industry or another bubble? Is it a defining trend or hype? – Learn how to separate the real deal from the nonsense and understand the game-changing AIaaS revolution and emerging commercial industries.

Supercharging Cutting-Edge Technology

While buzzwords come and go quicker than we can blink, actual game changers frequently emerge from the obscure world of defence research and development (R&D). To stay ahead of the curve, one should closely monitor the activities of defence agencies. Take DARPA's Information Innovation Office (I2O) for instance. People who paid attention could have seen the generative large language model (Gen LLM) craze coming in 2023 and the machine learning (ML) revolution coming in 2012. And this isn't simply a wild guess; it's about understanding how defence R&D influences the future of research and the technology industry.

Consider Google, Microsoft, OpenAI, and Anthropic getting their hands on DARPA's AI and cybersecurity initiatives, to mention a few. It's like being given the keys to the cutting-edge technology kingdom! This collaboration provides the ultimate boost, allowing for the rapid evolution of ground-breaking innovations.

But defence agencies don't just stop at collaboration; they take it to the next level by sharing algorithms and tools. It's like they're handing out cheat codes to speed up the development of experimental technologies.

This move is wise, as it not only accelerates innovation but also creates entirely new commercial industries based on these technologies, as we have seen before: microchips, GPS, the Internet, robotics, and much more.

Eight F**g Trends: Surfing the Wave of Defence Innovation

1. Emerging non-synthetic data platforms will play a crucial role in the AI revolution by enabling the mining, exchange, and commercialization of real-world data. Obviously, the capacity to seamlessly access, process, and extract value from huge datasets will be critical to being competitive in the AI race. This is especially important for autonomous AI.
2. Machine learning models (MLMs) and LLMs as a service: These may look like another bubble, but by using the power of Gen LLMs for data generation, they are gaining traction faster and faster, becoming a sweeping change of reality. For example, by evolving to data science as a service and killing data-related jobs, not to mention Gen LLM as a service.
3. Generative LLM (GeLLM) as a service: Three core trends are disrupting the tech world faster and faster–Gen LLMs as a service, AI Agents (mono products) built on Gen LLMs as a service, and High-Precision Gen LLMs as a service aimed at servicing Gen LLMs. The latter will specialise in accurate next-token predictions. These as-a-Service solutions are like having a tech genie in your pocket, granting wishes for innovative applications and unparalleled accuracy.
4. Convergence between MLMs, LLMs, and GeLLMs: The convergence of MLMs, LLMs, and GeLLMs is creating a behavioural insights boom that is revolutionising AI interactions further. This is about supercharging the development of highly personalised, engaging, and intuitive applications with understanding user behaviour deeper and deeper. It's like a mind-reading machine anticipating users' needs, mixing the predictive power of MLMs, contextual understanding of LLMs, and human-like responses of GeLLMs.
5. Open-source GeLLMs: The rise of open-source GeLLMs is another game-changer unleashed from the confines of proprietary silos. These models are giving developers and researchers unprecedented access to experimentation. Having just started, as more and more GeLLMs are open-sourced, we can expect to see an explosion of new applications and use cases that will transform industries and reshape the way we live and work.
6. Cyber AI: GeLLMs pose a threat of autonomous cyberattacks that help satisfy the data hunger of modern MLMs, enabling and enhancing simulation capabilities. These may aid in uncovering cyber vulnerabilities and fostering the development of adversarial AI. Such non-ethical AI can insert noise patterns into sensor data to misclassify self-learning models, affect their speed limit, or employ adversarial patching. It can conduct autonomous cyberattacks, not to mention the emerging deep fake and synthetic media generation tools available in the deep net.
7. Big Data Processing: The power trio of MLMs, LLMs, and GeLLMs is being leveraged to supercharge existing data processing algorithms, optimise server performance, storage, and energy consumption. In the meantime, the rise of quantum computing will add a whole new dimension to the mix, promising to revolutionise big data processing with its mind-bending ability to crunch numbers at lightning speed. In contrast, other defence-driven R&D is exploring alternative approaches to the traditional large language one to unlock unprecedented levels of performance and scalability without the help of quantum computers.
8. Autonomous Technologies: Driver-as-a-Service and Pilot-as-a-Service applications are the vanguard of this revolution, transforming transportation by making the once-futuristic notion of self-driving vehicles and autonomous aircraft a concrete reality—the driverless taxi Cruise in San Francisco and the pilotless F-16 under DARPA’s programme, to name a few.

What Businesses Are Appearing: Don’t Lose

The convergence of these trends and technologies is causing a gold rush for both ambitious startups and established companies looking to stake their place in the growing tech landscape and disrupt traditional sectors. While startups can use non-synthetic data platforms to create innovative AI agents, mature organisations can heavily leverage MLMs, LLMs, and GeLLMs as-a-service to stay ahead of the competition. Such technological convergence enables companies to create highly customised and engaging applications, opening new revenue sources and turning up consumer experiences to eleven.

Open-source GeLLMs are levelling the playing field by giving startups access to cutting-edge AI capabilities and the opportunity to compete with the big ones.

While cyber AI spurs new opportunities to offer autonomous cybersecurity solutions and secure code, autonomous technologies enable startups and established players to reimagine transportation and logistics, creating new business models and cranking up operational efficiency to warp speed.

Explainable AI is another coming game-changer that's set to spawn a whole new industry being baked at DARPA. This isn’t just some academic exercise; it's bloody integral to critical decision-making, especially when it comes to policy-making and governing. The problem with modern statistical ML approaches is that they're as opaque as a black box, making it a headache to ensure information integrity. LLMs might spit out compelling answers, but they also create a whole new can of worms when it comes to providing true explanations—even humans struggle with introspection and make up their own explanations half the time. But the businesses that can crack the code on Explainable AI will be sitting pretty, providing the services that ensure AI-driven decisions are transparent, accountable, and trustworthy—the new gold standard in critical industries like finance, healthcare, defence, and government.

At the crossroads of AI, Autonomous, and Cybersecurity, Authentication Technologies are giving birth to another commercial industry. Against the rise of deep fake and synthetic media generation tools, they're whipping up algorithms that can sniff out fake news and particular LLMs used, and they've got DARPA backing them up.

The Media Forensic and Semantic Forensic Programmes, for example, are the cutting edge of the fight against digital deception. And the companies that can harness these technologies and transform them into real-world solutions? They'll be the ones rolling in the dough, like Open AI, Anthropic, Space X, Tesla, Intel, Microsoft, and many other industry leaders. From tools for verifying the authenticity of news articles to platforms for detecting deep fakes in real time and services for safeguarding the integrity of digital content.

Deep-Deep Tech: Who Stays Behind the Scenes

If you want to stay ahead of the game in the world of deep tech and emerging industries, you'd better keep your eyes peeled on the movers and shakers in the defence innovation space.

These agencies are the puppet masters pulling the strings behind the scenes, cooking up the most cutting-edge, mind-bending technologies that'll shape our future. They're the ones pushing the boundaries of what's possible, exploring uncharted territories in AI, autonomous systems, cybersecurity, and beyond. So, if you want to know the future, you'd better get acquainted with these deep tech masterminds. Because the technologies they're cooking up today will be tomorrow's game-changers.

1. DARPA, the Defense Advanced Research Projects Agency, USA
2. IDEaS, the Innovation for Defence Excellence and Security, Canada
3. ARIA, the Advanced Research and Invention Agency, the UK
4. DDR&D, the Directorate of Defense, Research & Development, aka Maf'at, Israel
5. DSTA, the Defence Science and Technology Agency, Singapore
6. JEDI, the Joint European Disruptive Initiative, aka the European ARPA, the EU
7. SPRIN-D, the Federal Agency for Disruptive Innovation, Germany
8. Innosuisse, the Swiss Innovation Agency, Switzerland
9. AID, the Defence Innovation Agency, France
10. Vinnova, Sweden's innovation agency

The article was originally published on https://DmytroShestakov.com ?