This article is a sum-up of my random thoughts / notes / conclusions (you name it) related to latest-advancements in the artificial intelligence space and are based on my discussions with pioneers in the field while hanging around Asia’s biggest AI-focused conference located in Singapore, called: SuperAI.

AI is changing paradigm of addressing problems.

Mark Souza, Corporate VP & CTO Microsoft shared an interesting viewpoint about genAI redefining the way {we will} >> we are addressing problems. Rather than developing the solutions front to back in its entirety, what becomes significant is analyzing a challenge and carefully describing the task that we want AI system to solve.

And with that, people who want to prepare for the novel technology and future should ask themselves a question:

Where could the computer help me do what I wanted to accomplish, where does it help me to do my work better?

Rather than shouting in fear:

Where would it replace me?

Thinking about the future in the context of genAI would make more sense if we fear less about being replaced and explore more ways around augmentation of human work to make it more efficient, more impactful and effortless.

What does this mean in the context of the future of AI technology development?

For sure: foundation models set themselves apart from what we had in ML before, in the sense that they become gen-purpose machinery (we don’t train a foundation model for a particular task, but we can try utilize it for arbitrary tasks afterwards).

In my humble opinion – it’s still a long route before us. Why?

Because solving a problems that AI systems have never seen before remains a true challenge. Going beyond that, far beyond that – still poses a challenge and and interesting area for AI tech future research development.

AI is advancing 2 fast for it to be suppressed

We’re witnessing the birth of a child with an entirely new fresh technology, and already we’re seeing the werewolves beginning to circle

– Snowden told the SuperAI event in Singapore via remote call on Wednesday, June 5th.
Ben Goertzel did a great moderation during his combo panel appearance with Edward Snowden glimpsed into the present and future stats of AI’s global development.

In the context of:

• United States working on their regulatory approach to AI after popular services launch such as ChatGPT laid bare the technology’s potential;
• The European Union passing a sweeping law that would put guardrails on tech;
• and China, were no company can produce an AI service without proper approvals

makes a serious concern about governments and corporations trying to wrest control of artificial intelligence, potentially rendering the fast-developing tech less useful.

Finding a balance between necessary regulation and freedom for companies to innovate seems to be a critical aspect. As European, I advocate that we can’t leave behind the component that can enable industry, the component that can make us more competitive, that could truly revolutionize, but also re-industrialize our market and help our organizations grow.

If “they will build a dam around us, we have to make sure it doesn’t stop our flow”

– Snowden said.

Legendary NSA Expert signed off urging the audience to not “be afraid to defy consensus” as it might get us in trouble, but we’ll be much closer to being right…

AI Agents economic of scale and transformative Service-as-a-Software approach.

Some of the hottest discussions around here at SuperAI are floating around: The future of launching AI Agents in the enterprise.

AI Agent innovation is set to mirror the familiar pattern of disruptive innovation observed repeatedly in the tech sector. As many technology enthusiasts are aware, disruptive innovation typically begins by addressing challenges for the least served segment of the market with a fresh technology or business model. With time, it progresses to cater to a broader market and diverse use-cases as the product performance enhances.

Following Aaron Levie CEO at Box:

(…)While we’re in the earliest innings of AI Agent development right now, where most AI Agents can only handle a small fraction of work that we throw at them, and with often limited quality and accuracy. This is a dynamic that will inevitably change over time, and likely quite rapidly. The important thing about disruptive innovation is to understand what vector of improvement you can expect to continue to see over time that will cause the “low end disruption” to improve in quality over time.

(…)Something that may not work today or seems too expensive to automate, may just be one performance improvement cycle away from becoming possible.

Now, if we finally get to the point that AI Agents will be monetized the way we monetize professional services today, we’ll finally get to witness the full swing from: Software-as-a-Service to Service-as-a-Software. Monetization of software based on AI Agents doing i.e. web scraping or design work for your business $300/hr for similar output, but 1000x faster – that’s a serious breakthrough and new land of opportunities!

Building AI Agents has different economic of scale than employing thousands humans for the same job. Scaling AI technology in multiply causing exponential growth of new business opportunities should always come with a great responsibility (social impact, energy consumption, data privacy, democratization and more), and we shouldn’t put it into the hands of just few big tech companies out there.
The first day of SuperAI is over. I can’t wait what’s new at day number 2!

Ps Thank you SuperAI team for being good to me and welcoming so gentle at Marina Bay Sands Expo & Convention Center in Singapore! I can’t wait to join in 2025!