Browse Comments — Clean (de-noised)

Demis Hassabis As a software engineer who dreams of one day joining Google, I find these ambitious visions for AGI truly inspiring. What impresses me is not only the incredible technological advancements but also the simultaneous commitment to transparency, safety, and ethical frameworks, as exemplified by SynthID and CodeMender. Working in an environment that combines bold innovation with a responsibility to humanity is a dream come true. I hope to one day be part of this team that is writing the history of AI, not as a spectator but as a contributor to building the foundations of Singularity safely and effectively.

This is a major signal. What is emerging is not only a new generation of models, but a more integrated operating layer for AI: search, coding, science, content, agents, security, payments and user environments beginning to converge. That makes the governance question deeper than capability. As AI systems become more agentic and more embedded across everyday tools, institutions and users will need clear answers about data access, permissions, traceability, accountability and the ability to interrupt, contest or reverse action. The path toward more capable AI will also depend on whether governance can scale with the environments in which these systems operate.

I tested Gemini Omni Flash last night and I was truly impressed. Made a video of one of my PCs just by giving it a few pictures and asking it to combine everything into a smooth ad.

How do you know it's Information Theory??? 👀

Dr.Mohamed Nagy I do not think AI is a tool anymore because of the autonomy it now has

Did anyone mention how to secure all of these? Maybe run them by mythos 😁

Incredible pace of progress and some genuinely important breakthroughs especially around multimodal reasoning, agentic execution, scientific acceleration, and AI safety instrumentation. But the largest enterprise gap is no longer only model capability. The harder unsolved problems are operational governance, sovereign execution control, cross-agent state synchronization, runtime observability, operational memory consistency, real-time workflow orchestration and trusted enterprise execution boundaries That is where the industry still lacks mature operational foundations. Models are rapidly becoming more capable. Enterprise operational coherence is not scaling at the same pace. This is precisely where AI-native operational systems like MonkDB can play a major role by acting as the continuously synchronized operational intelligence layer across agents, workflows, telemetry, governance, memory, and enterprise execution systems.

The Fragmental Overlap Storage System

Thank your for making all Googlers proud of being here

What feels especially important is that AI progress is no longer advancing along a single axis of model capability. The frontier is now moving across multiple layers simultaneously: multimodal world understanding agentic coordination scientific discovery security infrastructure workflow integration institutional adoption That changes the nature of the race entirely. The organizations that shape the next era likely will not just build the most capable models. They will build the systems that organizations, governments, and individuals can reliably integrate into real-world decision-making over long periods of time. Capability matters. But trust, interoperability, coordination, and societal absorbability may matter even more.

I’m amazed by everything you’ve done, including your well deserved Nobel Prize, and have followed your podcasts/interviews since AlphaGo. There is genuine optimism in your vision of an “age of abundance.” But should we worry about the transition path? History suggests major tech revolutions create enormous wealth concentration before benefits spread. The Industrial Revolution improved billions of lives — but only after decades of inequality and social upheaval. This AI wave feels like tech change on chip steroids — both in scale and speed. Even today, we already see concentration of value creation around a relatively small number of companies, countries and ecosystems. If AGI becomes “10x bigger and 10x faster” than previous tech revolutions, the risk may not only be job displacement, but widening inequality — within countries, between capital and labor, and between advanced AI nations and the developing world. Will the “age of abundance” become genuinely inclusive — or risk becoming a mirage where much of humanity cannot participate in its benefits? Perhaps the challenge of this century is not whether we can build AGI, but whether we can ensure its benefits are shared broadly — while limiting misuse by bad actors.

Powerful progress — especially around agents, world understanding, CodeMender, and SynthID. The next frontier is not only what AI can understand or generate, but what it is allowed to execute. As agentic systems move closer to AGI, safety cannot remain only at the model or output layer. It needs a non-bypassable execution boundary: authorization, context, sensitivity, human approval where required, auditability, and real-time governance before action is released. In EMMTM terms: Capability can scale fast. Governance must scale deeper. Execution must remain bounded. That is where responsible AGI becomes structurally possible.

Exciting developments, especially around multimodal reasoning, scientific acceleration, and secure coding assistance. The pace at which AI capabilities are evolving is genuinely remarkable, and it’s encouraging to see strong focus areas like SynthID and CodeMender being treated as foundational rather than optional. At the same time, I think one of the biggest challenges ahead is ensuring organizations don’t treat AI safety, governance, and resilience as separate conversations from innovation. As agentic systems become more autonomous and deeply integrated into workflows, AI assurance will need to become part of architecture, cybersecurity, compliance, software engineering, and even leadership decision-making by default — not as an afterthought. The technology is advancing exponentially. Human oversight, governance models, and institutional readiness now need to evolve at comparable speed.

Demis Hassabis very exciting !!

So grateful to live in an era where high quality research, educative and creative tools are made available for everyone who wants to contribute to constructive projects worldwide.

Demis and Google 📈💯

".... foothills of singularity.." - I hope and pray that anyone reading this post may pause at these words, take a breath, and appreciate the gravity of these profound words..

The shift from models → agents → autonomous systems is happening much faster than most people realize. Feels like we're watching the operating system of the next decade being built in real time.

What makes this moment historically important is that AI is no longer evolving as a tool layer, but as a cognitive operating layer inside human systems. That shift changes the architecture of decision-making itself. The real challenge may no longer be intelligence generation, but preserving human clarity, strategic coherence, and contextual judgment inside increasingly agentic environments. This is where Cognitive Drift becomes critical. Very important direction — especially the focus on world understanding, agents, and responsible deployment as foundational infrastructure for the emerging Industrial Intelligence era.

Reading posts like this genuinely feels like watching the future arrive in real time. The shift from “AI that responds” to “AI that understands the world, acts, reasons, edits, secures, researches, and collaborates” is happening unbelievably fast. And honestly, the part about standing in the “foothills of the singularity” doesn’t even sound exaggerated anymore. Also really glad to see safety, transparency, and systems like SynthID being treated as core infrastructure instead of afterthoughts. The next few years are going to redefine almost every industry.