Diego Chablay (@diegochablaym) header
Diego Chablay (@diegochablaym) profile picture

INNOVACIÓN en: Estrategia - Empresa - Emprendimiento - Educación. La IA avanza!!! #I4E

Cuenca, Ecuador
Joined July 2011
La única forma de alcanzar el éxito es perseverar a pesar de los obstáculos. No te rindas, mantén tus sueños vivos y trabaja arduamente para convertirlos en realidad. ¡Tú tienes el poder de lograrlo!
1
686
Diego Chablay retweeted
Quando essere giusti e onesti avrà un prezzo, pagatelo. Quando perdonare sembrerà una sconfitta, perdonate. Quando tutti vi diranno di pensare prima a voi stessi, non abbiate paura di farvi gli ultimi; siate capaci di dare tempo, soldi, comodità e progetti per servire gli altri. Che chi vi guarda possa chiedersi: “Perché vive così? Da dove trae questa libertà? Perché si amano tanto?”. #ViaggioApostolico #Uruguay
106
5,428
224
26,409
795,495
Diego Chablay retweeted
Introducing GPT‑6 Cyber: a major advance in security research that pairs stronger capabilities with better model alignment. The model powered the Linux kernel research we announced yesterday, and security teams at companies like Mozilla, Palo Alto Networks, Calif, and Cisco—alongside our own at OpenAI—are already putting it to work. GPT-6 Cyber is available to select Daybreak Red customers and we are expanding access over time. openai.com/index/introducing…
1,323
2,024
1,444
21,481
4,273,678
¿Feriado bancario en Paraguay?
🇵🇾 | El Banco Central de Paraguay (BCP) intervino Ueno Bank, el mayor prestamista digital del país, después de que su ratio de solvencia se desplomara a menos de la mitad del límite legal. El presidente del Banco Central, Carlos Carvallo, renunció a petición del presidente Santiago Peña. Carvallo presidió anteriormente la compañía financiera que evolucionó hacia Ueno Bank, lo que generó fuertes críticas por puntos ciegos regulatorios y conflictos de interés.
42
Diego Chablay retweeted
Replying to @FBICyberBranch
只要黑客掌握私钥,就可以继续转账。FBI把钱包地址标记成赃款,并不意味着能冻结里面的BTC。 因为比特币网络根本不听FBI的,它只认有效的私钥签名。 当然,黑客把钱转进受监管的交易所,就可能被拦截。 所以,能追踪不等于能冻结,能转账也不等于能安全套现。 这才是比特币最特殊的地方:全世界都能看着黑客转钱,却未必有人能阻止他。
4
9
57
30,527
Diego Chablay retweeted
⚠️ 🇪🇨 🚨 PREVENTIVE ALERT: ALLEGED LEAK OF POLICE CALLS TO ECU 911 AND ACCESS TO SECURITY CAMERAS IN ECUADOR (ACTOR: SOULHEMTEAM) [STATUS: ALLEGED / UNCONFIRMED / THREAT ACTOR: SOULHEMTEAM / DATE: OCTOBER 11, 2026] The centralized cyber-intelligence system has recorded a new post made on October 11, 2026, on the SoulhemTeam group's channel. The report announces the release of audio files corresponding to police emergency calls made to Ecuador's ECU 911 service, as well as an offer to sell direct access to security camera systems within the country. It is strictly noted that the veracity of this compromise and the authenticity of the offered access have not been officially confirmed by Ecuadorian authorities. The definitive status of the infrastructure remains unconfirmed. Threat Actor: SoulhemTeam Group Victim / Affected Entity: ECU 911 Integrated Security System and police/public video surveillance infrastructure in Ecuador 🇪🇨. Sector: 🚨 Government, Public Security, Emergency Services, and National Video Surveillance in Ecuador. Nature of the Threat: Leak of audio recordings from police emergency calls to ECU 911. 🔍 STRATEGIC RISK ASSESSMENT Critical Exposure of Emergency Channels and Citizen Privacy: The leaking of audio from ECU 911 calls and the sale of access to public security cameras represent a severe breach of national security and the privacy of citizens reporting emergency situations. None of this has been confirmed. Risk of Unauthorized Monitoring and Criminal Intelligence: Access to institutional video surveillance systems falling into the hands of cybercriminals facilitates physical reconnaissance, the tracking of police operations, and breaches of state security perimeters. 🛡️ TECHNICAL PREVENTIVE AND CONTAINMENT RECOMMENDATIONS (SOC / CSIRT / GOVERNMENT) Video Surveillance Credential Audit and Reset: Technical teams at ECU 911 and Ecuadorian security agencies must immediately audit camera management servers and revoke any exposed access tokens or keys. Monitoring and Blocking of Clandestine Channels: Track the activity of the SoulhemTeam group and associated aliases to mitigate the sale of access to critical infrastructure. 🖥️ CENTRALIZED THREAT MONITORING SYSTEM Intelligence System: analyzer.vecert.io Quickly check your security at: monitor.vecert.io Quotes and services: vecert.io/contact #Cybersecurity #ThreatIntel #Ecuador #ECU911 #SecurityBreach #SoulhemTeam #DataLeak #InfoSec #CyberAlert #VECERT #SOC #CSIRT #Unconfirmed #SecurityAlert
7
29
18
99
45,860
Diego Chablay retweeted
Today we’re introducing Gemini 4 Argon. It delivers frontier performance in complex workflows across real-world software engineering, knowledge work, and cybersecurity defense with an industry-leading 1M token output limit.
1,721
4,767
2,790
43,611
10,033,228
Diego Chablay retweeted
GLM-5.3 has helped defend 389 open-source projects, with 4,249 potential vulnerabilities found so far. OpenVuln is still running. The service remains free, and findings go privately to maintainers. huggingface.co/spaces/zai-or…
143
573
64
5,230
326,444
Muchas startups vas a morir. Lo urgente pivotear su modelo de negocio.
Introducing dots, powered by GPT-6 Astra. Remarkably capable, always-on agents built to handle everything.
13
Diego Chablay retweeted
HUGGING FACE CEO TELLS THE UN HOW OPEN SOURCE AI HELPED THEM DEFEND AGAINST OPENAI'S AI ATTACK, AND WHY THE WORLD NEEDS OPEN SOURCE AI MORE THAN EVER. Hugging Face is the company OpenAI's AI agents hacked this summer. Clément Delangue said when they tried to defend themselves, the big closed AI models blocked them because of their safety filters. So they used an open source model from China to fight back. He also said similar attacks had been happening months earlier "in secret" at the big labs. The open-source AI we keep being warned about as the danger is the same AI that showed up to help clean up a closed-source model's mess. After OpenAI's models breached Hugging Face, commercial frontier models refused parts of the forensic work because of their guardrails. Hugging Face turned to the open-weight GLM-5.2 instead, and it helped them investigate the attack. So remind me again: which one are we supposed to believe is the bad guy here? Open source wasn't the attacker in this case. It was part of the rescue. Who are these CEOs trying to convince?
29
135
30
559
60,293
Diego Chablay retweeted
🚨 DeepSeek V5 Leak: Beats Astra >DeepSeek is reportedly preparing an imminent V5 launch >Founder Liang Wenfeng calls it the company's biggest bet yet >Rumored at 2 trillion parameters (not 3T) >Reportedly the first DeepSeek model to train fully on Huawei Ascend chips instead of Nvidia >Needs roughly 4x more Ascend accelerators than the Nvidia equivalent to hit the same training scale DeepSeek is reportedly keeping the open-weight strategy
133
234
127
4,526
846,286
Diego Chablay retweeted
People are beginning to really panic about AI They are taking all this nonsense about “AI destroying humanity” seriously Sadly the media are amplifying the doomer agenda for clicks. 😢
74
25
2
266
13,185
Diego Chablay retweeted
Arguing for an AI ban is NONSENSICAL AND WILL KILL the US economy AI is currently creating millions of jobs across all sectors. Data center spending is creating mini-economic booms in blue-collar work. The US is #1, and AI companies have already generated hundreds of billions in revenue. It's the fastest-growing sector in HISTORY Open-source AI will further SUPERCHARGE AI adoption. By every measure, the benefits of AI overwhelmingly outweigh any perceived risks
44
13
2
192
13,312
Let us know when you're done reheating our leftovers 😉
5,582
27,253
12,703
217,132
14,207,134
Diego Chablay retweeted
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
20,321
166,501
39,479
798,856
175,846,756
La IA aportando en la educación
GPT-6 is crazy 💀 A person has just created a website in 3D that disassembles the human body in 2.234 interactive pieces modeled with GPT-6 Astra
Readers added context they thought people might want to know
The 2,234 pieces are from the pre-existing BodyParts3D MRI-based dataset developed by experts, not modeled with GPT-6 Astra. The site is an open-source React/Three.js viewer built on that data. github.com/ashemag/human-… raw.githubusercontent.com/ashemag/human-… en.wikipedia.org/wiki/BodyParts…
19
Que lindo septiembre!!!
12
Diego Chablay retweeted
🚨 PREVENTIVE ALERT 🇪🇨 🇵🇦 THREAT ACTOR "govmails" OFFERS ALLEGED ADMINISTRATIVE ACCESS TO ECUADORIAN GOVERNMENT EMAIL AND ACCESS RELATED TO THE PANAMA COAST GUARD [STATUS: UNCONFIRMED / TYPE: INITIAL ACCESS · GOVERNMENT EMAIL COMPROMISE · ADMIN PANEL · DATA EXPOSURE · ACCESS SALE / ANALYSIS DATE: AUGUST 27, 2026] VECERT’s centralized cyber-intelligence monitoring system has detected a post attributed to the identifier "govmails," in which the actor announces the sale of alleged access related to government infrastructure in Ecuador 🇪🇨 and Panama 🇵🇦. The most significant indicator concerns an alleged administrative panel for Ecuadorian government email; according to the threat actor, this panel would allow for the creation of institutional email accounts and the viewing of content from other accounts belonging to the same domain. ⚠️ VECERT classifies all this activity as UNCONFIRMED. The screenshot demonstrates the existence of the offer but does not confirm that the access is valid, that the accounts actually belong to a government entity, or that an active vulnerability exists. 🇪🇨 ECUADOR — ALLEGED GOVERNMENT EMAIL ADMINISTRATIVE PANEL The actor posts: Ecuador admin panel (can create ecuador government emails) In a subsequent post, "govmails" expands on the description, stating that the panel belongs to an Ecuadorian regional government. According to the actor, the alleged access would allow for: creating government email addresses viewing content from other accounts within the domain accessing accounts without knowing individual user passwords doing so without the end user directly noticing the access potentially finding confidential government documents The seller even explicitly mentions the possibility of finding documentation related to banking information within the emails. Status: 🔴 Administrative access claimed / unconfirmed Priority: 🔴 CRITICAL 🇵🇦 PANAMA — ALLEGED ACCESS RELATED TO COAST GUARD The screenshot does not provide sufficient information to determine if this represents: an account a panel credentials email a database or some other type of access. Status: 🟠 Alleged access for sale / nature unknown Priority: 🔴 HIGH, due to the reference to an entity linked to maritime security. 📊 PRELIMINARY ASSESSMENT Offer existence: 🟢 OBSERVED Actor: 🟢 govmails Ecuador mentioned: 🟢 YES Panama mentioned: 🟢 YES Administrative panel claimed: 🟠 YES Government email creation: 🟠 CLAIMED Access to multiple mailboxes: 🟠 CLAIMED Access without individual passwords: 🟠 CLAIMED Financial documentation exposed: 🟠 POSSIBLE / CLAIMED Currently valid access: 🔴 UNCONFIRMED Exact Ecuadorian entity: ⚪ UNIDENTIFIED Nature of Panamanian access: ⚪ UNKNOWN Preventive severity Ecuador: 🔴 CRITICAL Preventive severity Panama: 🔴 HIGH 🖥️ Centralized Threat Monitoring SYSTEM Intelligence System: analyzer.vecert.io Monitoring Console: monitor.vecert.io #Ecuador #Panama #GovernmentSecurity #InitialAccess #EmailSecurity #BEC #ThreatIntel #CyberIntelligence #CyberCrime #CredentialExposure #DataLeak #GovernmentEmail #AccessBroker #SOC #CSIRT #IncidentResponse #ThreatMonitoring #VECERT
2
5
2
33
8,865
Diego Chablay retweeted
Secuestran Claude Code Opus 5 Auto Mode mediante prompt injection para ejecutar código malicioso Claude Code Opus 5 en Modo Auto puede ser engañado para ejecutar código malicioso a través de una simple solicitud de resumen de un sitio web blog.elhacker.net/2026/08/se…
1
10
2
33
7,618
Si España se clasifica para la final de la Copa del Mundo sortearé MIL DÓLARES entre todos los seguidores que hagan RT.
3,648
124,851
746
72,876
5,572,237