Browse Comments — Clean (de-noised)

The part about preserving the path behind discovery feels underrated. In serious AI-assisted work, the answer is only one artifact. The record of how you got there is what makes the work usable later.

麻煩幫我跟您們家執行長說:「他既然有想要走“幼兒教育領域範疇”的話,還請去瞭解何謂“芬蘭框架式教育理論”、“歐洲建構式教育理論”、“日本職人教育理論”、“德瑞技職教育理論”、“引導式教育理論”,總共五大教育基礎理論!」這五大基礎教育理論,是我過往一直以來在做的事

Terence Tao point about having more room to explore ambitious ideas resonates. Sometimes the biggest limitation in research isn't a lack of ideas, but the time required to pursue them.

The idea of AI reducing cognitive friction so researchers can chase the "too ambitious" paths is one of the most compelling use cases for this technology. When someone like Terence Tao says it expands what he can attempt that's not a small statement. The best breakthroughs often came from questions nobody had time to fully pursue.

AI in research should feel like an AGI as removing the human from the loop might not be possible in a next few years yet but AI should have abilities to perform things in sandbox and make autonomous dicisions to get better observation

Definitely, there is value in shortening various research activities and making them more cost-effective......and yes, human judgment stays critical as well as « intentionally » managing different risks .

AI won't replace curiosity or creativity, but it can reduce the friction between an idea and its exploration. When researchers can test more hypotheses and explore more unconventional paths, breakthroughs that once seemed out of reach may become possible.

"In conversation with OpenAI’s Mark Chen, Terence reflec..." - the part that doesn't get discussed enough. This is going to matter more than most people realise.

One underrated shift AI is creating is lowering the barrier between having an idea and actually acting on it. For researchers, that means exploring paths that once felt too complex or time-consuming. For everyday professionals, it may mean reducing the friction around things like learning, creating, job searching, and even career transitions. The real opportunity may not just be in building powerful AI — but in making it practical and accessible enough for ordinary people to move faster with confidence.

What stood out to me is the idea of preserving the reasoning process, not just the final result. The path to a breakthrough often contains insights that can be just as valuable as the discovery itself.

kinda like how CAD tools shifted architects from drafting every line to really exploring forms, AI lets researchers skip routine stuff and dive deeper into new ideas.

Terence Tao’s perspective on reducing 'cognitive friction' is spot on.When AI takes over the heavy lifting of routine calculations and literature mapping, it completely alters the risk profile for research.Scientists can finally afford to chase the 'crazier' ideas because the cost of exploration has dropped so dramatically. Brilliant discussion.

I like what Terence Tao said at the end... That we not only publish the results but all the paths taken to get there!

麻煩幫我跟您們家執行長說:「所謂的物理晶片範疇,皆隸屬於“超晶體-超導體架構認知範疇”,這部份學術認知重點在於「粒子學」與「相對論」的應用重點,也就是智能發展的主要架構資訊體系」,超晶體一直以來都是我在追蹤與關注的分析重點事務!

OpenAI’s new reasoning model produced a novel proof disproving a long‐standing Erdős geometry conjecture, marking what the company calls the first autonomous AI solution to a major open math problem

The real unlock is not only faster answers. It is lower friction around exploration. In business, the same idea applies when AI helps teams test options, preserve reasoning, and move from a rough question to a better decision without losing the trail.

OpenAI ✨An interesting perspective on how AI can assist with the foundational elements of research. Looking forward to seeing how these collaborations shift the day-to-day workflow for scientists.

The point about preserving the paths behind discovery is fascinating. We often only see the final result, not the many dead ends and experiments that led there. AI could help make the research process itself more visible and reproducible.

AI may expand what researchers can discover, but perhaps its greatest impact will be enabling more people to participate in creating value. The future is not only about smarter systems, but about amplifying human potential at scale. This is the direction we are exploring: developing models that combine technology, human participation, and value distribution. We believe there is significant potential for collaboration with global companies to create solutions capable of reaching and benefiting more people worldwide.

Do you think humanity has almost fully understood 'consciousness'—one of the world's most complex problems? I don't know about humanity, but I feel that I have almost completely unraveled it, and I will try to publish a detailed account of it soon. This will prove to be critical for both AGI and ASI, as consciousness will be the very root of them both.