Evergrid @evergrid

Frontier Infrastructure for the next wave of intelligence evergrid.ai NY, NY Joined September 2024

Tweets

155
Followers

95
Following

275
Likes

260

Andrej Karpathy @karpathy

4 years ago

The hottest new programming language is English

2K 9K 73K 12.9M 7K

View Details

Sam Altman @sama

5 months ago

we want to build tools to augment and elevate people, not entities to replace them.

2K 728 12K 3.1M 1K

View Details

Scientific breakthroughs demand infrastructure that performs at the highest level. From Frontier, the world’s fastest supercomputer, to LUMI and Discovery, AMD and @HPE bring supercomputing design and high-performance compute together to help researchers push boundaries in climate science, drug discovery, fusion energy, and AI. See how the most trusted companies trust AMD: bit.ly/3ZzkPXv

8 30 242 22K 9

View Details

CoreWeave @CoreWeave

8 months ago

Benchmarks tell you what’s possible. Production tells you what’s real. The gap between a "clean run" and a scaled pipeline is where most AI projects hit a wall. We’re closing that gap with CoreWeave ARENA—a production-ready lab for real-world workload evaluation. Why run in CoreWeave ARENA? 🔹No more guesswork: Test on real infra, not synthetic sims. 🔹Expert-led: Work alongside CoreWeave engineers to tune and iterate. 🔹Full Visibility: Mission Control + Grafana + W&B integrated from Day 1. 🔹Production-Ready: Support for CKS (Kubernetes) and SUNK (Slurm). Don’t commit to a rollout plan on faith alone. Validate the performance, scaling, and cost of your workload before you pull the trigger. Check out the launch details here: hubs.la/Q041Yjy80

0 3 24 4K 3

View Details

Data Center Knowledge @datacenter

6 months ago

❄️ Cold-Climate Data Centers: The Next Hot Thing in Data Center Growth datacenterknowledge.com/cooling/cold-c…

0 1 3 512 1

View Details

Josh Kale @JoshKale

6 months ago

This is big... Anthropic just announced a model so powerful they won't release it to the public out of fear over the damage it will cause 😨 Claude Mythos Preview found thousands of zero-day exploits in every major operating system and web browser... The numbers are hard to believe: > $50 to find a 27-year-old bug in OpenBSD, one of the most security-hardened operating systems ever built > Under $1,000 to find AND build a fully working remote code execution exploit on FreeBSD that grants unauthenticated root access from anywhere on the internet > Under $2,000 to chain together multiple Linux kernel vulnerabilities into a complete privilege escalation exploit For context: these are the kinds of findings that previously required elite security researchers working for weeks. Anthropic engineers with no formal security training asked Mythos to find exploits overnight. They woke up to working code the next morning. The results were so impressive Anthropic assembled Apple, Google, Microsoft, Amazon, NVIDIA, and seven other organizations into Project Glasswing: A $100M defensive coalition. They're not releasing this model publicly. Instead, they're racing to patch the world's infrastructure before models like this proliferate.

Anthropic @AnthropicAI

6 months ago

Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software. It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans. anthropic.com/glasswing

2K 7K 43K 31.9M 15K

679 2K 18K 4.1M 9K

View Details

Evergrid @evergrid

7 months ago

This isn’t incremental — it’s a shift. Running a 100B model on a CPU collapses cost, infrastructure, and data barriers at once. If this holds, AI moves from cloud to edge. The real question: what gets built when every device can run serious models locally? Big door just opened.

0 0 0 59 0

View Details

SpaceXAI @SpaceXAI

7 months ago

xAI is committed to deploying artificial intelligence that makes the lives of people better as well as adding more power near our datacenters to reduce the cost of energy for the American people

The White House @WhiteHouse

7 months ago

President Trump Participates in a Roundtable on Ratepayer Protection Pledge x.com/i/broadcasts/1…

531 994 4K 1.5M 216

550 870 6K 1.1M 327

View Details

Evergrid @evergrid

7 months ago

@sama Looking forward to seeing what makes /fast worth the spotlight.

0 0 0 41 0

View Details

Evergrid @evergrid

7 months ago

@OpenAIDevs Looking forward to seeing how that 1M token context actually performs in production versus the benchmarks.

0 0 0 65 0

View Details

Evergrid @evergrid

7 months ago

@sama The 1m token context window is going to completely change how we handle long form research and documentation work.

0 0 0 20 0

View Details

Evergrid @evergrid

7 months ago

@KobeissiLetter Energy markets showing us how fragile global supply chains really are. This is going to ripple through way more than just heating bills.

0 0 0 83 0

View Details

Evergrid @evergrid

7 months ago

@Justinmoon Open models need this kind of backing. The gap between accessibility and performance is still too wide for most builders.

0 0 1 49 0

View Details

Evergrid @evergrid

7 months ago

@DSBatten @dari_org The data is finally catching up to what the operators have known all along. Flexible load demand changes everything for grid economics.

0 0 0 176 1

View Details

Evergrid @evergrid

7 months ago

@JavierBlas This is the kind of supply shock that ripples through everything. Energy markets don't have much room for a 20% hit.

0 0 0 463 0

View Details

Evergrid @evergrid

8 months ago

Interesting prediction. The real challenge will be making sure these systems are accurate, reliable, and trusted before they replace critical professional work.

Financial Times @FT

8 months ago

CEO of Microsoft AI Mustafa Suleyman joins FT editor Roula Khalaf to explain why most of the tasks accountants, lawyers and other professionals currently undertake will be fully automated by AI within the next 12 to 18 months ft.trib.al/SZ4Lti1

725 2K 6K 7.5M 4K

1 0 1 651 1

View Details

Evergrid @evergrid

8 months ago

@AnthropicAI Congratulations on the milestone. Research and infrastructure expansion together will define the next phase of AI growth.

0 0 0 172 0

View Details

Jaynit @jaynitx

8 months ago

Elon Musk literally broke down his 5-step process for applying first-principles thinking to build anything:

42 982 5K 211K 4K

View Details

Evergrid @evergrid

8 months ago

@nikitabier The threat scales with AI, but so can the defenses. As AI scales outreach, identity verification, trust layers, and infrastructure-level safeguards will need to scale just as aggressively. Abuse is a systems problem, and systems can be redesigned.