←
AI67,196 posts
AI avatar

AI

@DeepLearn007

Imtiaz Adam CS #AI Postgrad |#Strategy #MachineLearning RSI Evolutionary Neurosymbolic |#RL #Agentic | #LLM Liberal | MBA alum @morganstanley @LBS @Columbia_Biz

◷ Joined September 2012▧ 6,309 Photos and videos
Nancy Pelosi Stock Tracker ♟ avatar
⟳ AI retweeted
Nancy Pelosi Stock Tracker ♟@pelositracker·22h

Something strange is happening 11/12 - outage at Microsoft 11/13 - US flights grounded nationwide 11/13 - outage payroll systems 11/13 - outage at Claude 11/16 - major US banks go down 11/17 - outage at NSA’s classified internal network Officials say none were cybersecurity attacks but the FBI warned that hackers can break into systems in under 3 hours

The #FBI issued a FLASH advisory warning that cybercriminals and state-aligned hackers are exploiting ARTEX, a super intelligence-powered penetration testing framework, to automate attacks against businesses and organizations worldwide. ARTEX can accelerate intrusions from initial targeting to privileged access in under three hours. It was used in the October ConnectWise supply-chain compromise, which exposed Microsoft 365 customer environments and involved attempted authentication against more than 2,250 downstream tenants. The FBI is urging businesses and organizations to strengthen access controls, monitor suspicious automated activity, and report potential compromises to ic3.gov. Read more here: ic3.gov/CSA/2026/261113.pdf

Post attachment
Tom Harwood avatar
⟳ AI retweeted
Tom Harwood@tomhfh·20h

Since Friday: 1.⁠ ⁠Microsoft down. 2.⁠ ⁠All US flights grounded. 3.⁠ ⁠⁠Major US bank service disruptions. 4.⁠ ⁠⁠NSA internal classified network down. 5.⁠ ⁠⁠Apple outages. 6.⁠ ⁠Magic Circle law firm ⁠Clifford Chance hit by major cyber attack. 7.⁠ ⁠⁠A separate cyberattack hits 178 of the world's biggest companies. 8.⁠ ⁠⁠Google catches Russia committing first fully autonomous cyber attack in history. 9.⁠ ⁠⁠The FBI issues a warning that cybercriminals and state-aligned hackers are exploiting AI penetration testing framework ARTEX, to automate attacks against businesses and organisations worldwide. Not all of these are cyber attacks. But they all reveal the vulnerability of our interconnected systems.

AI avatar
AI@DeepLearn007·3h

My experience: complex task build with Fable 5.2, review code with Astra 6.2 Less complex task build with Opus 5.5, review with Sol 6.1 High to Max settings. Result = less likely to get technical debt after production launch.

I tested GPT-6.2 Astra on real work. 2 repos, 105 replanted bugs. Find and fix what you can. It's the new #1. The results: - GPT-6.2 Astra (max): 62.7 for $28.35 - GPT-6.2 Astra (xhigh): 60 for $23.53 - Fable 5.2 (max): 58 for $139.65 - Gemini 4 Carbon (high): 57.3 for $31.32 - Sonnet 5.5 (max): 51.3 for $154.03 - GPT-6 Astra (max): 45 for $33.03 From GPT-6 Astra to GPT-6.2 Astra: +17.7 points at max, +17 at xhigh. Whatever people say about other use cases, it's a code review beast. GPT-6.2 Sol and more effort levels drop soon in the thread 🧵

Post attachment
Andrew Curran avatar
⟳ AI retweeted
Andrew Curran@AndrewCurran_·21h

Epoch says GPT-6.2 Astra continues the trendline above the old-world reasoning era, supporting my argument that we have been on a new trajectory since September, which will eventually become visible and undeniable.

Post attachment

GPT-6.2 Astra scores 174 on the Epoch Capabilities Index (ECI), on par with Fable 5.2 and joining it above what the pre-Astra trend for reasoning models would have predicted. This is based on testing we did prior to launch using only our own benchmarks. The ECI is continually updated as more benchmark results become available, and so these values may change in the future.

Post attachment
Tibo avatar
⟳ AI retweeted
Tibo@thsottiaux·10h

One of the funny reasons for the difference in vibes between the influencers and the crowd today on 6.2 Sol/Astra is that we gave a few folks access a while ago and then made a lot of changes in the last 3-4 days and I'm pretty sure no re-testing was done within those days and the videos were produced last week. Don't get me wrong, while the models are a hefty jump compared to their GPT predecessors, they are by no means perfect and there are a few things the teams are excited for the models to get better at in future iterations in terms of taste and niceness of use. Love all the feedback and we hear you!

Wevolver avatar
⟳ AI retweeted
Wevolver@WevolverApp·Nov 17

Researchers have used reinforcement learning to teach a 5-tonne John Deere skid-steer loader to perform complex acrobatic manoeuvres autonomously. Trained in simulation and transferred to a real machine, the system can balance on two wheels, navigate rough terrain, spin, and control its bucket before landing safely. The research demonstrates the potential of AI-driven automation for heavy construction and industrial machinery. Video Credit: @leggedrobotics / @ETH Paper: wevlv.co/4kltYys #engineering #technology

1:10
Demis Hassabis avatar
⟳ AI retweeted
Demis Hassabis@demishassabis·Nov 17

Thrilled to announce $5B in funding for @IsomorphicLabs to accelerate our mission to Solve All Disease. We continue to make rapid progress reimagining drug design from first principles with AI to radically cut the time & cost to find new medicines. Gradually, then all at once...

0:37
Watcher Guru avatar
⟳ AI retweeted
Watcher Guru@WatcherGuru·Nov 17

JUST IN: $13 trillion asset manager Charles Schwab's website goes down.

Timurs Jeparskis avatar
⟳ AI retweeted
Timurs Jeparskis@TJeparskis·18h

After testing GPT-6.2 Astra, I can say it’s definitely nowhere near as bad as people are making it out to be! The improvement over GPT-6 Astra is huge, and the model is now much more cost-efficient. I still think Fable 5.2 is better, but considering how much cheaper Astra is per task, it could be a really solid alternative!

Decado avatar
⟳ AI retweeted
Decado@ItsDecado·20h

They could hear a sixteen-year-old girl begging for her life. Somehow, the scandal was still who she loved. Sogand Sayadi lived in Kermanshah, Iran. In the account attributed to her classmates, someone reported that she was a lesbian and liked another girl at school. The religious woman in charge called her father. Sogand had warned that he would kill her. He beat her at school, in front of students and staff. She begged them not to let him take her home. They had the violence in front of their eyes and her warning in their ears. They let him take her. Her father is now accused of throwing her from a fifth-floor building. His name subsequently appeared on her funeral announcement among the grieving family. The man she feared was given a public place among those mourning her. Under the terrorist Islamic Regime occupying Iran, a woman can be forced to continue an unwanted pregnancy in the name of protecting life. Yet fathers who murder their children receive a legal exemption from the death penalty ordinarily applicable to murder. Protesters accused of burning trashcans have faced death sentences. Consider that hierarchy: a woman’s reproductive choices demand state intervention; a burning bin can justify an execution; murdering your daughter comes with a special provision for fatherhood. Apparently, the obligation to protect your child becomes an advantage in court when you kill her. Sogand needed someone to take her fear as seriously as they took her feelings for another girl. Spare me the fucking sermons about the sanctity of life. She asked them to protect hers.

Post attachment
Robin Ebers avatar
⟳ AI retweeted
Robin Ebers@robinebers·17h

had early access to the astra 6.2 and sol 6.2 for quite some time really no no comprehensive review because they won't replace opus for me they're exactly what you expect of OpenAI models - super technical - best for in-depth reviews - overengineers as per usual - relentless in its pursuit to succeed they're really NOT BAD models at all people telling you otherwise are baiting you the most interesting part is that the $200 plan now has Ultrafast, though this will burn your plan like there's no tomorrow in one breath: they won't change my workflow, but if all other models would disappear tomorrow, i'd be very grateful to work with these two /s

Derya Unutmaz, MD avatar
⟳ AI retweeted
Derya Unutmaz, MD@DeryaTR_·21h

I’ve had early access to both GPT-6.2 Astra & Sol, I can unequivocally say that both models are absolutely amazing! I’m incredibly impressed by how much they’ve improved over the 6.1 models! I’ll be sharing several eg soon, I think people will quickly realize how good they are!

Some of you have been asking very politely. Some of you have just been asking. GPT-6.2 Astra and Sol are now available. Super efficient and built to perform where it counts—the real world. openai.com/index/introducing…

1:59
Elshayib avatar
⟳ AI retweeted
Elshayib@elshayib_·20h

All the Anthropic shells are attacking OpenAI with full force at once I hate to defend OpenAI i don't like them either but I'm fair and I call out bullshit when I see it. Astra and Sol 6.2 are nowhere near as bad as they're making them out to be this is ridiculous why are they not talking about the INSANE pricing or speed improvement? This is coordinated just before the IPO

Conor Dart avatar
⟳ AI retweeted
Conor Dart@Conor_D_Dart·20h

I am testing GPT 6.2 Astra and Sol, and so far I don’t see anything negative. Ignore the negative reviews until YOU have tested it yourself

SkalskiP avatar
⟳ AI retweeted
SkalskiP@skalskip92·17h

GPT-6.2 is getting roasted on X meanwhile, Astra and Sol just took #1 and #2 on my object detection benchmark so I have some awkwardly positive results to share ↓ benchmarks and examples

Post attachment

I will not be writing a full review of GPT-6.2 Astra / Sol. I had a terrible experience testing them. They were essentially unusable. I do not recommend these models.

Epoch AI avatar
⟳ AI retweeted
Epoch AI@EpochAIResearch·21h

GPT-6.2 Astra scores 174 on the Epoch Capabilities Index (ECI), on par with Fable 5.2 and joining it above what the pre-Astra trend for reasoning models would have predicted. This is based on testing we did prior to launch using only our own benchmarks. The ECI is continually updated as more benchmark results become available, and so these values may change in the future.

Post attachment