Dmitrii Sherstobitov (@12bytez) header
Dmitrii Sherstobitov (@12bytez) profile picture

Database Staff

Sunnyvale, CA
Joined October 2015
Photo by Dmitrii Sherstobitov (@12bytez)Photo by Dmitrii Sherstobitov (@12bytez)Photo by Dmitrii Sherstobitov (@12bytez)Photo by Dmitrii Sherstobitov (@12bytez)
Dmitrii Sherstobitov retweeted
your agent after it's 50th compaction
19
197
14
3,734
68,205
Dmitrii Sherstobitov retweeted
Join us TOMORROW (Sept 25) at 9:30am PDT for a special engineering roundtable: Agentic Programming Practices at @Yugabyte - youtube.com/watch?v=PfKQCWNe… Panelists: Hideaki Kimura Dmitrii Sherstobitov Rajagopalan Madhavan Nikhil Chandrappa Hari Krishna Sunder Host: Me 😎 Five of our Yugabyte engineers share how agentic coding works day-to-day on a large distributed database codebase, including what each one tried, kept, and moved on from. We'll cover agentic code review, how much human review different kinds of work still need, bug hunting and production debugging with agents, and sharing project context across teammates and agents with Meko, our agent context engine. We'll also talk about what didn't work. These practices are still changing month to month, and the panel will explain why they made the choices they did.
1
1
5
554
Dmitrii Sherstobitov retweeted
Looked deeply into Laya. It's interesting. The timeline claims are fuzzy and benchmarks are mixed but the model is worth following. Laya, as compared to Jev by the author and @benchmarkheaven: -#4 overall w/ 70.1 vs Jev’s 75.4 -Much stronger cost score: 86 vs 52 -Weaker intelligence, calibration + speed -34% vs 74% on hard questions Most interesting result comes from Laya’s own 2,000-decision benchmark... When *fine-tuned* its typed-decisions checkpoint scored 0.766 vs Jev’s published 0.727. Laya trained on 6,000 decisions from the same 4 synthetic workflows, whereas Jev ran zero-shot. This is a narrow, same-distribution win, to be super clear. But ~0.36 base → 0.766 fine-tuned is a big jump. Expect us to go hard on fine-tuning open weight models. Lots of caveats on these benchmarks and more empirical methods are definitely needed. But closed weight models can't be fine-tuned. Whoever can create a highly fine-tunable Jev-like model will win the market.
7
7
56
4,214
Dmitrii Sherstobitov retweeted
We're adding support for AGENTS.md to Claude Code. Starting today in version 2.1.277, if there is no CLAUDE.md in a folder, Claude will check for and use AGENTS.md. You can toggle this behavior in /config.
2,048
2,625
2,523
31,270
5,644,066
Dmitrii Sherstobitov retweeted
Today, we’re announcing Ternary Bonsai 2 27B. Based on Qwen3.8 27B, Bonsai 2 27B is 9x smaller than its full-precision counterpart while retaining 98.2% of its aggregate benchmark performance. Two months after the first Bonsai 27B release, the biggest change is quality. The footprint remains 5.9 GB, but the gap to full precision has narrowed materially, with particularly strong gains in agentic coding, multimodal reasoning, and long-horizon tool use. Ternary Bonsai 2 27B is available today under Apache 2.0.
633
1,563
790
15,480
5,510,093
Dmitrii Sherstobitov retweeted
"we have sandboxed the agent" the agent:
139
515
152
6,214
6,791,260
Dmitrii Sherstobitov retweeted
As our CFO @_balaji_km mentioned at earnings today, we’re seeing some very interesting trends on AI costs. I think it’s another signal that we’re coming to the end of the so-called ‘tokenmaxxing’ era. Here’s what’s been happening behind the scenes. Since the beginning of the year we’ve more than quadrupled the number of people using frontier AI tools. That’s thousands of engineers using them every single day. During that same period, our cost per token has declined. You might expect costs to rise as adoption accelerates. We've seen the opposite. Not because we've restricted access, but because we've treated efficiency as an engineering problem rather than a budget problem. A few examples: • Caching and reuse: We use optimizations to improve our prompt cache hit rate that reduce our input token spend. • Better defaults and tooling: We tuned default model settings, context sizes and developer workflows so teams get the same results with fewer tokens and lower-cost inference. • Visibility drives efficiency: We gave engineers real-time visibility into their AI usage and costs per hour. • Experimenting with open-weight models: we continuously evaluate new models and deploy the best option for each use case. This is the future of applied AI at enterprise scale. The next phase, whatever we call it, will not be characterized by who spends the most tokens, but about how people use them as efficiently as possible. Credit to all the engineers at @Uber who are helping to build this future. 🚀
76
146
59
1,586
501,775
Dmitrii Sherstobitov retweeted
AI can polish the work. Can you defend the reasoning behind it? Caliber measures judgment, evidence, trade-offs and uncertainty in AI-assisted work.

2
3
2
43
180,919
Dmitrii Sherstobitov retweeted
The Sovereign NOC: A Hybrid Human-Agent Operations Case Study dimaggi.ai/insights/sovereig…
🤖 Made with AI
2
1
3
124
Dmitrii Sherstobitov retweeted
AI models write vanilla #PostgreSQL. If your database is distributed, providing the #AI model with a #YugabyteDB skill file closes the gap and ensures it writes code that works for your application.🔥 In his latest blog, @Yugabyte expert Dmitry Sherstobitov breaks down #benchmarking results across 17 model configurations after conducting 350+ evaluations across #Claude 4.5, 4.6, and 4.7, #Gemini 3.1 Pro, #GPT-5.x, #Composer 2, #CodexCLI - through #Claude Code CLI, #Cursor, and #Codex. 👀 Discover what moved the score, what regressed, and the one structural finding that changes how every team should think about delivering context to an AI coding tool!💡⬇️ na2.hubs.ly/H05DGVd0 @AnthropicAI, @GeminiApp, @OpenAI, @cursor_ai, @xai
2
2
208
Dmitrii Sherstobitov retweeted
I don't need Pomodoro anymore, my workday has natural breaks now
91
457
58
6,789
182,289
Dmitrii Sherstobitov retweeted
Had to go see Project Hail Mary right away (it's based on the book of Andy Weir, of also The Martian fame). Both very pleased and relieved to say that 1) the movie sticks very close to the book in both content and tone and 2) is really well executed. The book is one of my favorites when it comes to alien portrayals because a lot of thought was clearly given to the scientific details of an alternate biochemistry, evolutionary history, sensorium, psychology, language, tech tree, etc. It's different enough that it is highly creative and plausible, but also similar enough that you get a compelling story and one of the best bromances in fiction. Not to mention the other (single-cellular) aliens. I can count fictional portrayals of aliens of this depth on one hand. A lot of these aspects are briefly featured - if you read the book you'll spot them but if you haven't, the movie can't spend the time to do them justice. I'll say that the movie inches a little too much into the superhero movie tropes with the pacing, the quips, the Bathos and such for my taste, and we get a little bit less the grand of Interstellar and a little bit less of the science of The Martian, but I think it's ok considering the tone of the original content. And it does really well where it counts - on Rocky and the bromance. Thank you to the film crew for the gem!
353
314
94
9,050
735,712
Dmitrii Sherstobitov retweeted
Excited to launch FOUR new AI @Reddit communities designed to empower global and African innovation, research, ethics & entrepreneurship! Join us to connect, learn, and shape the future of AI. A thread... #AI #Africa #MachineLearning #GlobalAI #Artificial_Intelligence
1
2
4
116
Tech industry vet and CMU alum @mzweben today launches his new podcast, @MLMinutes, with interviews of @mldcmu's Tom Mitchell and Manuela Veloso, who also is head of AI research at @jpmorgan. What a great start! apple.co/3411DG8
2
8
5
16
Не знаю, кто автор, но это прям пять 😆🍷
23
1,058
68
6,940
Dmitrii Sherstobitov retweeted
Вся динамическая типизация в одной картинке nitter.cf/so_mad_so_funny/status…
3
18
Dmitrii Sherstobitov retweeted
Как завещал @varlamov : надо делать игровые площадки такими, чтобы было весело и взрослым – at Moscow, Russia
6
21
2
201
Dmitrii Sherstobitov retweeted
Ничего не меняется – at Saint Petersburg, Russia
6
137
10
540