Who's Who of AI

Sung Kim

244 trust @sungkim.bsky.social · 8,036 followers
A business analyst at heart who enjoys delving into AI, ML, data engineering, data science, data analytics, and modeling. My views are my own. You can also find me at threads: @sung.kim.mw
What they're sharing

Articles & links

Anthropic's When AI builds itself "We looked at sessions where a human researcher took a wrong turn, showed Claude the session up to that point, and asked it what to do next. Mythos Preview improved on humans 64% of the time—up from 22% in 2024." www.anthropic.com/institute/re...

When AI builds itself anthropic.com
View on Bluesky · ♥ 17 ↻ 2 ↩ 1 · 15 from the directory shared this · 14d ago
↻ Sung Kim reposted
@isolyth.dev

New gemma!!! And it's a diffusion model! Deepmind keeps releasing diffusion stuff 🤔 it's not that much worse on benches compared to the same sized autoregressive Gemma 4

DiffusionGemma: 4x faster text generation blog.google
AI Weekly's analysis →
  • DiffusionGemma generates 256 tokens per forward pass using bidirectional attention, reaching 1,000+ tokens/sec on a single H100 GPU.
  • With only 3.8B active parameters during inference and an 18GB VRAM footprint when quantized, it runs on consumer hardware without server-grade resources.
  • Google recommends DiffusionGemma only for speed-critical workloads like in-line editing and code infilling, not for applications requiring maximum quality.
Read full analysis →
 View on Bluesky →

www.ft.com/content/d02d...

ft.com
View on Bluesky · ♥ 8 ↻ 0 ↩ 0 · 4 from the directory shared this · 13d ago
Their own posts

Recent commentary

I call B.S. on this. Anthropic would cut them off before this happens. They want to get paid after all.

View on Bluesky · ♥ 50 ↻ 2 ↩ 7 · 20d ago

I'm guessing Anthropic is also working on their own custom AI chip.

View on Bluesky · ♥ 52 ↻ 1 ↩ 5 · 11d ago

I remember people cacelling their OpenAI subscription because they sold their AI to US Defense Department. Me: I’m not cancelling. They’re all same. It seems Anthropic is personally supporting NSA to spy on us. Like they have people at NSA onsite, helping them.

View on Bluesky · ♥ 38 ↻ 6 ↩ 4 · 13d ago

In the age of AI coding agents, can a single hybrid architect/developer outperform a traditional team of architects and engineers using the same tools? Funny enough, based on my experience, the answer is yes.

View on Bluesky · ♥ 37 ↻ 2 ↩ 8 · 13d ago

Nvidia’s Nemotron 3 Ultra. A 550B MoE frontier-intelligence open model built for long-running agents. It delivers 5x faster inference and lowers the cost of complex agentic tasks by up to 30% versus other open frontier models.

View on Bluesky · ♥ 40 ↻ 6 ↩ 2 · 14d ago

Anthropic: Use Fable/Mythos to harden your app against security threats. Me: Tries to use Fable/Mythos to harden my app against security threats. Anthropic: Fable 5's safety measures flagged this message for cybersecurity or biology topics. They may flag safe, normal content as well.

View on Bluesky · ♥ 33 ↻ 3 ↩ 5 · 9d ago

It seems Chinese LLMs are having a pricing war.

View on Bluesky · ♥ 39 ↻ 2 ↩ 2 · 25d ago

Let the sovereign AI war begin. For years, people have worried that countries would eventually restrict access to frontier AI in order to preserve national competitive advantage. The recent USG directive against Anthropic’s Fable 5 effectively codified that concern.

View on Bluesky · ♥ 37 ↻ 2 ↩ 2 · 5d ago

What I don’t understand is the assumption that human software developers are inherently better than AI coding agents. That may be true for the top tier of human developers, but for the vast majority, I don’t think the gap is as obvious as people want to believe.

View on Bluesky · ♥ 30 ↻ 2 ↩ 4 · 14d ago

Anthropic: Don't compete with us, using our tools.

View on Bluesky · ♥ 28 ↻ 3 ↩ 2 · 9d ago
Their network

In Sung Kim's orbit

Center = Sung Kim. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.