Browse Comments — Raw (as collected)
Close reading of the corpus at each pipeline stage: raw → clean → relevant → coded.
250
comments matched
· page 1 of 13
UniversalAssistant.ai !
Congrats Demis!!!
Gemini co - scientist is really impressive. Demis Hassabis.
What stands out is the gap between capability and controllability.
World understanding, code repair, scientific agents, multimodal generation, watermarking. All impressive. But once systems can reason across modalities and act across domains, safety has to move from “model behavior” to “consequence control.” Who authorized the action? What evidence was used? What changed? Can it be reversed? Who is accountable?
World class contribution to society
Impressive set of tools! Is there any work toward watermarking biotechnology or biomedical research?
Even along side all of these amazing achievements I still believe that coordination around the adoption of SynthID and similar watermarking modalities will yield the greatest amount of impact in the short run. Wishing you continued progress on this front.
"DeepMind focuses on connecting models like Genie (a world model) with SIMA (an agent) to create infinite training loops".
Is it possible to see the first effects in Gemini 3.5 Pro?
A compelling glimpse into where value creation is heading.
Agentic AI, multimodal generation, and domain-specific applications like CodeMender or Gemini for Science are not just technical milestones—they have the potential to reshape cost structures, accelerate R&D cycles, and redefine competitive advantage across sectors.
As we move closer to AGI, the winners will be those who can industrialise these capabilities responsibly and at scale.
Congratulations- it was a joy to hear your story.
Demis Hassabis The pace of AI advancement is becoming extraordinary.
What stands out most is that the conversation is no longer just about models — it’s about world understanding, agents, multimodal reasoning, scientific acceleration, and infrastructure-level integration across industries.
The shift from AI tools to truly agentic systems is happening faster than most businesses realize. Equally important is the focus on safety, transparency, and responsible deployment as capabilities continue scaling toward AGI.
Exciting — and historic — times for the technology ecosystem.
Good to here human touch by Google
The transition to AGI may not primarily change what systems can do.
It may change what humans are no longer required to perceive directly.
And that is where governance stops being a technical layer
and becomes a civilizational necessity.
•_•
The real breakthrough is not that models are becoming multimodal.
It is that they are gradually evolving from pattern-recognition systems into world-modeling systems.
The moment AI can continuously model reality, simulate consequences, and act across environments with persistent memory and reasoning, the economic and geopolitical implications become far larger than software itself.
At that point, AI stops being a tool layer.
It becomes infrastructure
Thank you for highlighting Gemini for science; this is such an important effort in human advancement!
Congratulations! Amazing models and improvements. 💯🚀🦾 #GoogleAI
I like that you have taken a multi-modality approach from the start.
AGI isn't the finish line, it's the starting gun for a completely different competitive game. The companies that figure out how to operationalize these capabilities at enterprise scale in the next 24 months are going to own the next decade.
Great share
Token Zombie Apocalypse
I am working on "Zero-waste token architecture" to save humanity from Token Zombies.
In a multi-agent AI company, token waste can grow quickly because many agents may:
- Talk too much.
- Repeat research.
- Re-read the same files.
- Store too much memory.
- Retrieve irrelevant context.
- Use expensive models for simple work.
- Spawn unnecessary agents.
- Reflect too often.
- Debate decisions that do not matter.
- Send full histories instead of compact summaries.
- Use LLMs for tasks deterministic code can solve.
www.threatclaw.ai
www.xodexa.com
#ZeroWasteArchitecture #TokenOptimization #MultiAgentSystems #AgenticAI #LLMOps #AIOrchestration