Ray
@300signs.bsky.social
Masonry dreams of architecture, and architecture must remember the stone.
@300signs (here)
@500signs
@160signs (on X)
Tech Innovator.
@300signs (here)
@500signs
@160signs (on X)
Tech Innovator.
Pinned
Ray
@300signs.bsky.social
· 10d
🇪🇺 Happy 2026 Welcome to all who share our freedom and values 🇪🇺 the rest may stay home !
Sam raises important points on AI capability, cost, and resilience. Yet one critical layer remains underexplored: information safety — a public-health approach to the ecosystem of ideas (public data | communications), analogous to food safety for physical health
eu.36kr.com/en/p/3657136...
eu.36kr.com/en/p/3657136...
Altman Admits "Messed Up": GPT - 5.2 Sacrifices Writing for Programming, 100x Cost Reduction Next Year, Agents Confirmed for Permanent Work
How to Integrate AI into Early Childhood and Basic Education?
eu.36kr.com
January 28, 2026 at 6:33 AM
Sam raises important points on AI capability, cost, and resilience. Yet one critical layer remains underexplored: information safety — a public-health approach to the ecosystem of ideas (public data | communications), analogous to food safety for physical health
eu.36kr.com/en/p/3657136...
eu.36kr.com/en/p/3657136...
🇪🇺 Happy 2026 Welcome to all who share our freedom and values 🇪🇺 the rest may stay home !
January 22, 2026 at 7:31 AM
🇪🇺 Happy 2026 Welcome to all who share our freedom and values 🇪🇺 the rest may stay home !
Anti‑woke absolutism and moral‑reasoning AI cannot fully coexist. In a complex world, refusing regulation is not neutrality but an abdication of reason, and morally reasoning AI makes that abdication increasingly untenable.
techcrunch.com/2026/01/21/a... via @techcrunch
techcrunch.com/2026/01/21/a... via @techcrunch
Anthropic revises Claude's 'Constitution,' and hints at chatbot consciousness | TechCrunch
The newly revised document offers a roadmap for what Anthropic says is a safer and more helpful chatbot experience.
techcrunch.com
January 22, 2026 at 7:26 AM
Anti‑woke absolutism and moral‑reasoning AI cannot fully coexist. In a complex world, refusing regulation is not neutrality but an abdication of reason, and morally reasoning AI makes that abdication increasingly untenable.
techcrunch.com/2026/01/21/a... via @techcrunch
techcrunch.com/2026/01/21/a... via @techcrunch
W Social will Twitter Konkurrenz machen www.bilanz.ch/people/w-soc...
W Social setzt auf echte Nutzer und Werte
Ein neues soziales Netzwerk soll Europa unabhängiger machen. In der Hauptrolle: Die Zürcherin Anna Zeiter.
www.bilanz.ch
January 21, 2026 at 8:07 AM
W Social will Twitter Konkurrenz machen www.bilanz.ch/people/w-soc...
Today’s AI models are far more capable than how we use them.
It’s like driving a sports car on narrow city roads: the power is there, but the infrastructure, tools, and habits keep us slow.
The real challenge now isn’t smarter AI — it’s better ways for humans to use i
eu.36kr.com/en/p/3635615...
It’s like driving a sports car on narrow city roads: the power is there, but the infrastructure, tools, and habits keep us slow.
The real challenge now isn’t smarter AI — it’s better ways for humans to use i
eu.36kr.com/en/p/3635615...
GPT - 5.2 Surpasses Humans in Exams: OpenAI Warns of Excessive Large - Model Capabilities and Non - AI AGI Ceiling
Similar to smartphones, large models have stepped into an era of "excess capabilities," with a significant disparity existing between the inherent capabilities of large models and the utilization meth...
eu.36kr.com
January 12, 2026 at 5:52 AM
Today’s AI models are far more capable than how we use them.
It’s like driving a sports car on narrow city roads: the power is there, but the infrastructure, tools, and habits keep us slow.
The real challenge now isn’t smarter AI — it’s better ways for humans to use i
eu.36kr.com/en/p/3635615...
It’s like driving a sports car on narrow city roads: the power is there, but the infrastructure, tools, and habits keep us slow.
The real challenge now isn’t smarter AI — it’s better ways for humans to use i
eu.36kr.com/en/p/3635615...
Ask our Canadian friends to translate
January 10, 2026 at 12:59 PM
Ask our Canadian friends to translate
When AI turns hiring into a black box, people bypass it. Some now use dating apps to find jobs. That’s not innovation — it’s a system failure.
gizmodo.com/ai-ruined-jo...
gizmodo.com/ai-ruined-jo...
AI Ruined Job Applications, So People Are Resorting to Dating Apps to Find Work
Networking has become the only way to get a job.
gizmodo.com
January 5, 2026 at 7:43 AM
When AI turns hiring into a black box, people bypass it. Some now use dating apps to find jobs. That’s not innovation — it’s a system failure.
gizmodo.com/ai-ruined-jo...
gizmodo.com/ai-ruined-jo...
US's 2025 straightjacket
December 18, 2025 at 9:52 AM
US's 2025 straightjacket
If a last-minute code red can turbo-boost a whale like ChatGPT, the lab army clearly isn’t bound by human rules. ChatGPT isn’t OpenAI’s product anymore — it’s their first and most demanding client. AI: turning products into self-users.
www.techbuzz.ai/articles/ope...
www.techbuzz.ai/articles/ope...
OpenAI rushes GPT-5.2 launch to Dec 9th as 'code red' hits
Exclusive: OpenAI accelerates GPT-5.2 release to counter Google's Gemini 3 dominance
www.techbuzz.ai
December 6, 2025 at 5:54 AM
If a last-minute code red can turbo-boost a whale like ChatGPT, the lab army clearly isn’t bound by human rules. ChatGPT isn’t OpenAI’s product anymore — it’s their first and most demanding client. AI: turning products into self-users.
www.techbuzz.ai/articles/ope...
www.techbuzz.ai/articles/ope...
"war has been declared" - effective from Jan. 15th 2026 -
Meta is Removing Microsoft Copilot and ChatGPT From WhatsApp: Key Takeaways For IT Leaders www.uctoday.com/unified-comm...
Meta is Removing Microsoft Copilot and ChatGPT From WhatsApp: Key Takeaways For IT Leaders www.uctoday.com/unified-comm...
Meta is Removing Microsoft Copilot and ChatGPT From WhatsApp: Key Takeaways For IT Leaders
Meta's new WhatsApp terms force ChatGPT and Copilot off the platform by Jan 2026. Here's what IT leaders need to know about the change.
www.uctoday.com
November 30, 2025 at 4:46 PM
"war has been declared" - effective from Jan. 15th 2026 -
Meta is Removing Microsoft Copilot and ChatGPT From WhatsApp: Key Takeaways For IT Leaders www.uctoday.com/unified-comm...
Meta is Removing Microsoft Copilot and ChatGPT From WhatsApp: Key Takeaways For IT Leaders www.uctoday.com/unified-comm...
To our kids' teachers
openai.com/index/chatgp...
openai.com/index/chatgp...
A free version of ChatGPT built for teachers
A secure ChatGPT workspace that supports teachers in their everyday work so they can focus on what matters most—plus admin controls for school and district leaders. Free for verified U.S. K–12 educato...
openai.com
November 30, 2025 at 12:20 PM
To our kids' teachers
openai.com/index/chatgp...
openai.com/index/chatgp...
this morning 😅:
<me>: refine the following text "Ray is not an X supporter"
<copilot>: I'm sorry, but it seems I can't help out with this one.
<me>: refine the following text "Ray is not an X supporter"
<copilot>: I'm sorry, but it seems I can't help out with this one.
November 28, 2025 at 6:04 AM
this morning 😅:
<me>: refine the following text "Ray is not an X supporter"
<copilot>: I'm sorry, but it seems I can't help out with this one.
<me>: refine the following text "Ray is not an X supporter"
<copilot>: I'm sorry, but it seems I can't help out with this one.
You don’t need to be LGBTQ to feel distant from your biological family. Most of my life was shaped by my chosen family. Family is who you build, not necessary who you’re born with.
www.rd.com/article/chos...
www.rd.com/article/chos...
November 26, 2025 at 7:20 AM
You don’t need to be LGBTQ to feel distant from your biological family. Most of my life was shaped by my chosen family. Family is who you build, not necessary who you’re born with.
www.rd.com/article/chos...
www.rd.com/article/chos...
Am@zing tool ! edisonscientific.com/articles/ann...
Kosmos: An AI Scientist for Autonomous Discovery
Today, we are announcing Kosmos, our next-generation AI Scientist. Kosmos is a major upgrade on Robin, our previous AI Scientist. You can read about it in our technical report, here. Kosmos is availab...
edisonscientific.com
November 16, 2025 at 10:50 PM
Am@zing tool ! edisonscientific.com/articles/ann...
China’s ambitions stay modest because its “vassal” adds little: 140M people, poor, aggressive, and contributing almost nothing to global progress except perhaps in its thirst for being in a state of perpetual war. Russia’s place in today’s world is hugely overstated — SOS fraud!
November 16, 2025 at 5:26 AM
China’s ambitions stay modest because its “vassal” adds little: 140M people, poor, aggressive, and contributing almost nothing to global progress except perhaps in its thirst for being in a state of perpetual war. Russia’s place in today’s world is hugely overstated — SOS fraud!
In relativity, space and time are unified into spacetime, whose geometry is governed by Lorentz transformations—but you can grasp the concept even without math.
bigthink.com/starts-with-...
bigthink.com/starts-with-...
How to understand Einstein's relativity without math
Times dilate and lengths contract near the speed of light. Bizarre and confusing? Sure. But under relativity, it can't be any other way.
bigthink.com
November 7, 2025 at 8:41 AM
In relativity, space and time are unified into spacetime, whose geometry is governed by Lorentz transformations—but you can grasp the concept even without math.
bigthink.com/starts-with-...
bigthink.com/starts-with-...
Too bad ChatGPT Atlas is launching on 🍎 first — an all-in-one Chromium-based browser with search and AI chat is exactly what we’ve been waiting for www.tomsguide.com/news/live/op...
ChatGPT Atlas launch live — OpenAI takes on Chrome with new browser
A ChatGPT internet browser is here
www.tomsguide.com
October 22, 2025 at 5:49 AM
Too bad ChatGPT Atlas is launching on 🍎 first — an all-in-one Chromium-based browser with search and AI chat is exactly what we’ve been waiting for www.tomsguide.com/news/live/op...
Sorry for being proud. Turns out, I actually have a reason this time 😁
techcrunch.com/2025/10/14/y...
techcrunch.com/2025/10/14/y...
You'll soon be able to shop Walmart from ChatGPT | TechCrunch
Walmart is partnering with OpenAI to let shoppers buy products directly through ChatGPT. The new integration will allow users to link their Walmart accounts, browse items, and instantly check out — al...
techcrunch.com
October 15, 2025 at 8:04 AM
Sorry for being proud. Turns out, I actually have a reason this time 😁
techcrunch.com/2025/10/14/y...
techcrunch.com/2025/10/14/y...
About gap between training data and release - Realistic expectation:
If GPT-6 releases around mid-2026, its knowledge base will probably run up to Q1 2026, possibly March 2026 or later - much closer to real time than any previous version (18 months for v3.5, 11 months for v4, 6-8 months for v5).
If GPT-6 releases around mid-2026, its knowledge base will probably run up to Q1 2026, possibly March 2026 or later - much closer to real time than any previous version (18 months for v3.5, 11 months for v4, 6-8 months for v5).
October 14, 2025 at 5:20 AM
About gap between training data and release - Realistic expectation:
If GPT-6 releases around mid-2026, its knowledge base will probably run up to Q1 2026, possibly March 2026 or later - much closer to real time than any previous version (18 months for v3.5, 11 months for v4, 6-8 months for v5).
If GPT-6 releases around mid-2026, its knowledge base will probably run up to Q1 2026, possibly March 2026 or later - much closer to real time than any previous version (18 months for v3.5, 11 months for v4, 6-8 months for v5).
to keep an 👀 on
techcrunch.com/2025/10/07/h...
techcrunch.com/2025/10/07/h...
How Otter.ai's CEO is pushing the company to be more than just a meeting scribe | TechCrunch
Otter is launching a suite of enterprise tools to help companies use its tech to create a central knowledge base.
techcrunch.com
October 8, 2025 at 4:20 AM
to keep an 👀 on
techcrunch.com/2025/10/07/h...
techcrunch.com/2025/10/07/h...
I’m not angry to be copied by OpenAI 😄 — if anything, I can benefit from their ecosystem and especially from their monetization ideas to strengthen my upcoming family of solutions -
OpenAI launches apps inside of ChatGPT techcrunch.com/2025/10/06/o...
OpenAI launches apps inside of ChatGPT techcrunch.com/2025/10/06/o...
OpenAI launches apps inside of ChatGPT | TechCrunch
OpenAI is launching a new way for developers to build applications that appear directly inside of ChatGPT's conversations.
techcrunch.com
October 7, 2025 at 6:58 AM
I’m not angry to be copied by OpenAI 😄 — if anything, I can benefit from their ecosystem and especially from their monetization ideas to strengthen my upcoming family of solutions -
OpenAI launches apps inside of ChatGPT techcrunch.com/2025/10/06/o...
OpenAI launches apps inside of ChatGPT techcrunch.com/2025/10/06/o...
I usually don’t need an external keyboard for my computers, but this one really impresses me
www.logitech.com/en-us/shop/p...
www.logitech.com/en-us/shop/p...
Logitech Signature Slim Solar+ Keyboard for Mac, Solar and Artificial Light Powered
A wireless keyboard designed for Mac powered by renewable solar energy and available artificial light, with a modern and sleek design which helps you manage your blended life in a zero-hassle way.
www.logitech.com
October 6, 2025 at 8:25 AM
I usually don’t need an external keyboard for my computers, but this one really impresses me
www.logitech.com/en-us/shop/p...
www.logitech.com/en-us/shop/p...
A friend asked me why I'm neglecting LinkedIn a bit
September 18, 2025 at 5:23 PM
A friend asked me why I'm neglecting LinkedIn a bit
When we expect instant breakthroughs, we act like the very systems we critique — mechanical, impatient, blind to nuance. True intelligence, human or artificial, requires humility, context, and time.
quillette.com/2025/09/07/c...
quillette.com/2025/09/07/c...
ChatGPT-5 and the Limits of Machine Intelligence
The disillusion produced by GPT-5 is not a technical hiccup, it’s a philosophical wake-up call.
quillette.com
September 8, 2025 at 6:32 AM
When we expect instant breakthroughs, we act like the very systems we critique — mechanical, impatient, blind to nuance. True intelligence, human or artificial, requires humility, context, and time.
quillette.com/2025/09/07/c...
quillette.com/2025/09/07/c...
We’re flirting with disaster if we continue to believe that AI truly embodies the kind of intelligence we associate with human thought. What we’re dealing with—and likely will be for quite some time—is not intelligence in that deeper sense, but something better described as <<Anthropic Computing>>
July 30, 2025 at 1:26 PM
We’re flirting with disaster if we continue to believe that AI truly embodies the kind of intelligence we associate with human thought. What we’re dealing with—and likely will be for quite some time—is not intelligence in that deeper sense, but something better described as <<Anthropic Computing>>