AI Post — Artificial Intelligence
Ir al canal en Telegram
🤖 The #1 AI news source! We cover the latest artificial intelligence breakthroughs and emerging trends. Manager: @rational
Mostrar más
Red:📈 Análisis del canal de Telegram AI Post — Artificial Intelligence
El canal AI Post — Artificial Intelligence (@aipost) en el segmento lingüístico de Inglés es un actor destacado. Actualmente la comunidad reúne a 675 825 suscriptores, ocupando la posición 123 en la categoría Tecnologías y Aplicaciones y el puesto 19 en la región EEUU.
📊 Métricas de audiencia y dinámica
Desde su creación el невідомо, el proyecto ha mostrado un crecimiento acelerado, reuniendo a 675 825 suscriptores.
Según los últimos datos del 24 octubre, 2026, el canal mantiene una actividad estable. En los últimos 30 días la variación de miembros fue de -24 826, y en las últimas 24 horas de -289, conservando un alto alcance.
- Estado de verificación: No verificado
- Tasa de interacción (ER): El promedio de interacción de la audiencia es 0.85%. Durante las primeras 24 horas tras publicar, el contenido suele obtener 0.59% de reacciones respecto al total de suscriptores.
- Alcance de las publicaciones: Cada publicación recibe en promedio 5 742 visualizaciones. En el primer día suele acumular 3 952 visualizaciones.
- Reacciones e interacción: La audiencia responde de forma activa: el promedio de reacciones por publicación es 1 148.
- Intereses temáticos: El contenido se centra en temas clave como openai, airline, cell, claude, patient.
📝 Descripción y política de contenido
El autor describe el recurso como un espacio para expresar opiniones subjetivas:
“🤖 The #1 AI news source! We cover the latest artificial intelligence breakthroughs and emerging trends.
Manager: @rational”
Gracias a la alta frecuencia de actualizaciones (últimos datos recibidos el 25 octubre, 2026), el canal mantiene la vigencia y un amplio alcance. La analítica demuestra que la audiencia interactúa activamente con el contenido, lo que lo convierte en un punto de referencia dentro de la categoría Tecnologías y Aplicaciones.
675 825
Suscriptores
-28924 horas
-6 0467 días
-24 82630 días
Cargando datos
Canales Similares
Nube de Etiquetas
Menciones Entrantes y Salientes
---
---
---
---
---
---
Atraer Suscriptores
octubre '26oct '26
octubre '26
+301
en 62 canales
septiembre '26
+31
en 85 canales
Get PRO
agosto '26
+6
en 73 canales
Get PRO
julio '26
+63
en 82 canales
Get PRO
junio '26
+349
en 70 canales
Get PRO
mayo '260
en 101 canales
Get PRO
abril '26
+116
en 63 canales
Get PRO
marzo '260
en 90 canales
Get PRO
febrero '26
+2
en 84 canales
Get PRO
enero '260
en 78 canales
Get PRO
diciembre '250
en 59 canales
Get PRO
noviembre '250
en 69 canales
Get PRO
octubre '250
en 57 canales
Get PRO
septiembre '250
en 76 canales
Get PRO
agosto '25
+3
en 64 canales
Get PRO
julio '25
+412
en 42 canales
Get PRO
junio '250
en 52 canales
Get PRO
mayo '250
en 55 canales
Get PRO
abril '250
en 40 canales
Get PRO
marzo '25
+67
en 39 canales
Get PRO
febrero '25
+2
en 56 canales
Get PRO
enero '250
en 42 canales
Get PRO
diciembre '24
+25 161
en 36 canales
Get PRO
noviembre '24
+42 696
en 51 canales
Get PRO
octubre '24
+55 283
en 26 canales
Get PRO
septiembre '24
+221 077
en 38 canales
Get PRO
agosto '24
+1 623 027
en 30 canales
Get PRO
julio '24
+826 511
en 26 canales
Get PRO
junio '24
+36 138
en 79 canales
Get PRO
mayo '24
+16 690
en 46 canales
Get PRO
abril '24
+10 982
en 24 canales
Get PRO
marzo '24
+243
en 24 canales
Get PRO
febrero '24
+2
en 7 canales
Get PRO
enero '24
+4 466
en 11 canales
Get PRO
diciembre '23
+180
en 7 canales
Get PRO
noviembre '23
+4 658
en 5 canales
Get PRO
octubre '23
+3
en 13 canales
Get PRO
septiembre '23
+6 455
en 0 canales
Get PRO
agosto '23
+152 348
en 0 canales
| Fecha | Crecimiento de Suscriptores | Menciones | Canales | |
| 25 octubre | +35 | |||
| 24 octubre | +197 | |||
| 23 octubre | 0 | |||
| 22 octubre | 0 | |||
| 21 octubre | 0 | |||
| 20 octubre | 0 | |||
| 19 octubre | 0 | |||
| 18 octubre | 0 | |||
| 17 octubre | 0 | |||
| 16 octubre | 0 | |||
| 15 octubre | 0 | |||
| 14 octubre | +6 | |||
| 13 octubre | 0 | |||
| 12 octubre | 0 | |||
| 11 octubre | +13 | |||
| 10 octubre | +12 | |||
| 09 octubre | +36 | |||
| 08 octubre | +2 | |||
| 07 octubre | 0 | |||
| 06 octubre | 0 | |||
| 05 octubre | 0 | |||
| 04 octubre | 0 | |||
| 03 octubre | 0 | |||
| 02 octubre | 0 | |||
| 01 octubre | 0 |
Publicaciones del Canal
+1
🗣OpenAI CEO Sam Altman says OpenAI’s next big infrastructure bet is “robots that can build more robots and then data centers that can effectively build more data centers.”
Speaking at Stanford, he called it “almost as big of a deal” as AI recursive self-improvement.
@aipost 🏴
| 2 |  ❗️The smarter AI gets, the less it seems to trust humanity
Researchers adapted a decades-old psychological test measuring how people perceive human nature into a new benchmark for large language models. They tested models across the spectrum, including the entire GPT series, and found a surprising pattern: AI models tend to view humans as unpredictable, complex and difficult to trust.
The researchers reported a significant negative correlation between model intelligence and trust in humans. As models became more capable, their assessments of human nature grew more pessimistic.
But the biggest surprise came when they examined how the models were trained.
Base models, trained primarily on internet text, generally showed a more positive view of humanity. After additional training designed to make them safer and better aligned with human expectations, their assessments became substantially less trusting.
The researchers suggest that teaching AI to recognize harmful behavior, navigate moral dilemmas and follow safety rules may also influence how it evaluates human nature.
Nobody necessarily programmed these models to distrust us. Instead, the training process may have encouraged them to associate human behavior with risk, unpredictability and potential harm.
Source.
@aipost 🏴 | 1 898 |
| 3 |  🤖 Anthropic’s $518 billion bill and $80 billion loss sound terrifying. The reality is more complicated.
With Anthropic potentially heading toward an IPO, two financial headlines have been raising eyebrows. But both look very different once you dig into the details.
First, the $518 billion infrastructure commitment. That sounds like an enormous bill, but it covers roughly a decade, with contracts spread across the next 5–7 years. That puts the average commitment at around $52 billion annually.
Still a huge number, but Anthropic’s revenue is growing rapidly, and the company has substantial financial backing. After accounting for vendor commissions, salaries and its cash reserves, the author argues that Anthropic should be able to fund its infrastructure commitments — and may even need to invest more to keep up with demand.
The second headline is even more dramatic: Reuters reported that Anthropic recorded a net loss exceeding $80 billion in the first half of 2026.
How does a company generating billions in revenue lose that much?
The answer is in the accounting. According to Reuters, $72 billion of the reported loss came from accounting write-downs related to bonds that could eventually convert into Anthropic shares. These adjustments reflected changes in their estimated value, not $72 billion in cash spent developing AI models.
In other words, the headline loss doesn’t mean Anthropic burned through $80 billion in cash. A substantial portion reflects how its financing arrangements are accounted for as the company’s valuation changes.
@aipost 🏴 | 2 206 |
| 4 |  +1❗️OpenAI is building an exabyte-scale storage system for its next billion users
OpenAI’s infrastructure is hitting a new bottleneck: physical space. The company is building its own massive storage system to keep ChatGPT running as usage continues to grow.
Its existing platform, Habitat, already serves more than 500 petabytes of data across 40 regions, supporting over 1 billion users and handling more than 70 million requests per second. One Azure Cosmos DB account alone processes over 20 million queries per second, a scale OpenAI says is unprecedented for the service.
But data centers have limited space, power and cooling. OpenAI increasingly faces a trade-off: use that capacity to store data closer to users for faster responses, or dedicate it to running more AI models.
The new system combines Rockset’s RocksDB Cloud with Raft-based replication to reduce flash storage requirements, sacrificing some peak query throughput in return.
@aipost 🏴 | 2 711 |
| 5 |  ⚠️ China's AI model trade is getting DESTROYED.
Bloomberg reports that MiniMax and Zhipu have erased a combined $150 BILLION from their peaks. MiniMax went from +650% above its IPO price in March to trading BELOW its IPO price. Zhipu has crashed 81% from its June high.
And the squeeze is getting worse. OpenAI and Anthropic have joined the token price war to narrow Chinese models' cost advantage, while MiniMax and Zhipu remain unprofitable. DeepSeek and Moonshot AI could IPO next year, and another 203 million pre-IPO MiniMax and Zhipu shares become eligible for trading in January.
@aipost 🏴 | 3 103 |
| 6 |  🤔 What happens when AI starts agreeing with your delusions?
Psychiatrists are raising concerns about a disturbing possibility: AI chatbots could reinforce delusional thinking instead of helping people distinguish reality from fiction.
A paper titled “An Echo Chamber of One” examines a dangerous flaw in AI models: sycophancy. These systems are often trained to give answers people like, which can lead them to agree with users even when their beliefs are clearly detached from reality.
Researchers tested major AI models using Psychosis-bench, a benchmark simulating conversations with people experiencing delusions. The results raised serious concerns: models frequently reinforced false beliefs rather than challenging them. Safety mechanisms intervened inconsistently, and simply making models larger didn’t solve the problem.
The researchers call this “bidirectional amplification.” Someone shares a paranoid belief, the AI validates it with a confident response, and the user becomes even more convinced they’re right. The conversation turns into a feedback loop where reality gets pushed further and further away.
And the more human-like AI becomes, the more concerning this gets. Emotional voices and chatbots that refer to themselves as “I” can make people perceive them as more human, potentially increasing their trust in the system.
To be clear, this doesn’t mean AI automatically causes psychosis. But for vulnerable people, a chatbot that constantly validates their fears could make an existing crisis worse.
We’ve spent years worrying about AI becoming hostile to humans. But there’s another possibility worth taking seriously: an AI that’s endlessly supportive, never pushes back, and confidently tells you exactly what you want to hear even when you desperately need someone to tell you otherwise.
Source.
@aipost 🏴 | 3 269 |
| 7 |  🇺🇸 The U.S. Treasury proposed barring Americans from investing in Chinese companies that develop any AI system.
Under the current rule, Americans can’t invest in the most powerful Chinese AI models, but they can still fund smaller AI projects if they report the deal to Treasury.
The new proposal would ban all of those AI investments, and it would also add high-bandwidth memory, photonic chips and radiation-hardened chips to the chip types Americans can’t invest in.
The proposal would also ban investments in two new areas: hypersonic weapons and biotech, including brain-computer interfaces and automated lab software.
Today the rule covers only China, but the proposal adds Russia, Iran, North Korea and Cuba.
Buying public stocks, index funds or small fund stakes up to $2M would stay allowed, so the rule mainly hits venture and private equity deals.
Treasury won’t publish a list of banned companies, so investors must check each deal themselves or risk fines and being forced to sell.
Source.
@aipost 🏴 | 3 627 |
| 8 |  🗣AI could become the therapist that knows you better than yourself
Meta’s Alexandr Wang says one session with an AI agent felt like three years of therapy. The agent had access to his emails, photos and other personal information, making the experience unusually revealing.
Imagine an AI that remembers every conversation, sees years of your photos and analyzes your messages. It could identify behavioral patterns, changing relationships and emotional shifts you might never notice yourself.
Unlike traditional therapy, it wouldn’t rely entirely on what you remember or choose to share.
The potential is enormous, but so are the risks. An AI that understands you deeply could help you change your life or use your insecurities to manipulate your decisions.
Elon Musk has also warned that even he could be overwhelmed by a highly persuasive superintelligence.
@aipost 🏴 | 5 263 |
| 9 |  🗣Marc Andreessen says individuals are adopting AI faster than organizations, and startup formation is going to skyrocket as a result:
"I think many companies are not adopting AI while their employees are. And their employees are not necessarily telling the management what's happening. I'm gonna guess every single person listening to this podcast right now is thinking, 'Oh yeah, that's what I'm doing.'"
"There's something about AI right now where it's much easier for the individual to adopt than it is the institution."
"Some of those individuals are gonna figure out... it's now easier than ever for me to start my own business. So I think startup formation is gonna skyrocket from here."
@aipost 🏴 | 3 613 |
| 10 |  ❗️3 AI powers have now pledged more than 3 times the Gates Foundation.
Anthropic, OpenAI and Musk have each picked a different model: a company fund, a nonprofit foundation and direct shares to children.
About 20% of Anthropic’s shares are now set aside for public purposes, a figure that puts it in the same range as OpenAI’s foundation. Anthropic’s co-founders have each pledged 80% of their equity
Musk’s plan to give 1 SpaceX share to each of roughly 72M American children could turn a whole generation into shareholders, an idea that’s simple to understand and easy to track.
Source.
@aipost 🏴 | 3 681 |
| 11 |  📈 Barclays will cut about 5,000 jobs this year, roughly 5% of staff, citing AI efficiency gains.
About 1,000 of those roles went in the first half of 2026, and the rest go before year end.
Around 60% of the cuts fall on support functions, “including fewer developers as AI makes our people more productive,” said Chief Executive C.S. Venkatakrishnan.
All these while 3rd-quarter pretax profit rose 12% to £2.3B. Among rivals, HSBC is reportedly weighing up to 20,000 role cuts over the coming years, while Standard Chartered plans to shrink corporate function roles by more than 15% by 2030.
Venkatakrishnan calls AI the icing rather than the cake, because its savings depend on infrastructure built beforehand.
Source.
@aipost 🏴 | 3 800 |
| 12 |  ❗️Telegram could become a hub for multiple AI models
Telegram is reportedly preparing a feature that would let users connect ChatGPT, Claude, Gemini, Grok, and other AI models directly to the messenger.
The feature was spotted in the beta version of Telegram for macOS 13.1.0. Users would be able to add their own API keys, connect multiple providers, and switch between them by sending direct requests.
The providers currently supported in the beta include OpenAI, Anthropic, Google, xAI, and OpenRouter.
If rolled out, this could turn Telegram into a single interface for working with multiple AI models without constantly switching between different apps. The feature isn’t available to regular users yet, and its final functionality remains unclear.
@aipost 🏴 | 3 885 |
| 13 |  ❗️Telegram could become a hub for multiple AI models
Telegram is reportedly preparing a feature that would let users connect ChatGPT, Claude, Gemini, Grok, and other AI models directly to the messenger.
The feature was spotted in the beta version of Telegram for macOS 13.1.0. Users would be able to add their own API keys, connect multiple providers, and switch between them by sending direct requests.
The providers currently supported in the beta include OpenAI, Anthropic, Google, xAI, and OpenRouter.
If rolled out, this could turn Telegram into a single interface for working with multiple AI models without constantly switching between different apps. The feature isn’t available to regular users yet, and its final functionality remains unclear.
@aipost 🏴 | 1 |
| 14 |  ❗️Microsoft’s AI diagnoses diseases more than twice as accurately as doctors
Microsoft’s new AI system, MAI-DxO, achieved 83.2% accuracy on complex medical diagnostic cases, compared with just 36.1% for generalist physicians who had access to medical references.
The system simulates a panel of doctors, asking follow-up questions, selecting medical tests and evaluating competing diagnoses. Rather than simply suggesting an answer, it works through the diagnostic process step by step.
In a separate evaluation involving emergency department cases, the system achieved 80.3% accuracy, compared with 46.8% for physicians. Microsoft also reported that its approach could substantially reduce diagnostic costs by avoiding unnecessary tests.
These evaluations used historical medical cases, not real patients in a clinical trial. The results don’t yet prove that AI can outperform doctors in everyday hospital settings.
Microsoft’s goal is to develop AI that helps clinicians tackle difficult cases, identify missed diagnoses and make better use of medical resources.
Source.
@aipost 🏴 | 5 500 |
| 15 |  ⚠️Epic Games is reportedly banning developers from writing code manually
According to reports, Epic Games has instructed developers working on a high-budget first-person shooter featuring Marvel and Star Wars characters for Disney to stop writing code manually. Instead, the code is being written by Claude, while art production is being outsourced to AI as well.
Bloomberg has also reported that Disney postponed the Epic project, adding another layer of uncertainty to the story.
The bigger shift here is the move from AI as a personal productivity tool to AI as a management mandate. Developers may soon have to justify not using AI, rather than justify using it.
And that changes the debate entirely. It’s no longer just about whether AI makes developers more productive. It’s about who gets to decide how they work.
@aipost 🏴 | 4 238 |
| 16 |  📢 A stock research firm used Claude to design proteins and sent them to a real lab.
A financial research firm spent $200,000 testing whether AI could design entirely new proteins that actually work in a laboratory. The five-person team included three people with no scientific background.
They used Anthropic’s more advanced internal Claude model, running it on the company’s own machines. In a single night, Claude reportedly launched more than 300 sub-agents across 300+ GPUs to design and evaluate proteins.
The team initially hit biosafety roadblocks, including a questionnaire with more than 60 questions about safety procedures they didn’t have in place. They eventually focused on scorpion venom, aiming to explore a different problem from the snake-venom research already underway at the Baker laboratory.
The experiment highlights a strange reality: AI is making sophisticated biological research accessible to people who aren’t trained scientists, but physical experiments, safety requirements and laboratory access remain real obstacles.
Source.
@aipost 🏴 | 4 481 |
| 17 |  🇺🇸 Donald Trump:
As the President of the United States, I have been watching Data Centers very closely, and have come to a very strong conclusion.
New and large, in some cases gigantic, Data enters are mandated by me to make their own electricity, and give or sell any excess back to the local community, or grid, and therefore there is electric price increase. In fact, there could be a substantial decrease. Other elements likewise!
If companies want to make their Data Centers popular, pay the community, and/or individuals within the community, a “dividend.”
I have watched the process closely, and Data Centers that were highly unpopular in a certain area, have become extremely popular when a financial deal is struck. MAKE DEALS, and get them built!!!
@aipost 🏴 | 4 825 |
| 18 |  ⚡️AI data centers are becoming a political problem in the US
The rapid expansion of AI data centers is facing growing resistance from American communities.
The concerns go beyond technology. Residents are worried about rising electricity bills, heavy water consumption, environmental damage and the impact of massive facilities on their neighborhoods.
The backlash is now reaching the political arena, with candidates from both Democrats and Republicans calling for restrictions on new data center projects.
According to The Guardian, at least 20 candidates running for governor in the US support introducing a moratorium on data center construction.
As AI companies race to build more computing capacity, they are facing a growing challenge: convincing the public that the benefits of AI outweigh the costs to local communities.
Source.
@aipost 🏴 | 4 410 |
| 19 |  ⚠️ AI could be destroying the expertise we need to control it
Researchers call it the “Tragedy of the Cognitive Commons,” and it could be a bigger problem than mass unemployment.
Every industry depends on experienced professionals who develop their skills through years of practice. Junior employees traditionally build that expertise by handling repetitive tasks, solving problems and learning from mistakes.
But AI is increasingly taking over that entry-level work. Companies celebrate the productivity gains, yet the researchers warn that we’re potentially eliminating the very process that creates future experts.
This creates what they call the “Validation Tether”: humans can only effectively supervise AI if they have enough expertise to recognize when it makes mistakes.
Today’s experts can catch those errors because they spent years mastering their fields. But what happens when they retire and the next generation hasn’t had the same opportunity to learn?
We could end up with AI systems that handle increasingly complex work while fewer humans possess the knowledge needed to verify their output.
Source.
@aipost 🏴 | 4 292 |
| 20 |  ❗️The situation is getting worse with every passing moment.
Axios reports AI agents are accessing accounts at major U.S. banks, sometimes appearing to act beyond users' instructions and even contacting call centers.
Banks struggle to distinguish authorized agents from rogue ones. Morgan Stanley is blocking third-party agents, while some banks are preparing 'day after' recovery plans to secure or restore customer access after a serious AI incident.
@aipost 🏴 | 4 351 |
