Vladimir Pikalov retweeted
i can't believe nobody caught this.
Anthropic's entire growth marketing team was just ONE PERSON
(for 10 months, confirmed)
a single non-technical person ran paid search, paid social, app stores, email marketing, and SEO for the $380B company behind claude
here's exactly how one human is doing the job of a full marketing team:
it starts with a CSV.
1. he exports all his existing ads from his ad platforms along with their performance metrics (click-through rates, conversions, spend, etc)
2. feeds the whole file into claude code
3. and tells it to find what's underperforming.
claude analyzes the data, flags the weak ads, and generates new copy variations on the spot
this is where he gets clever:
he then splits the work into 2 specialized sub-agents:
1. one that only writes headlines (capped at 30 characters)
2. and one that only writes descriptions (capped at 90 characters).
each agent is tuned to its specific constraint so the quality is way higher than cramming both into a single prompt
so now he's got hundreds of fresh headlines and descriptions.
but that's just the text.
he still needs the actual visual ad creative, the images and banners that go on facebook, google, etc.
so he built a figma plugin that:
1. takes all those new headlines and descriptions
2. finds the ad templates in his figma files
3. and automatically swaps the copy into each one.
up to 100 ready-to-publish ad variations generated at half a second per batch.
what used to take hours of duplicating frames and copy-pasting text by hand
so now the ads are live.
the next question is which ones are actually working.
for that he built an MCP server (basically a custom integration that lets claude talk directly to external tools) connected to the meta ads API.
so he can ask claude things like:
⢠"which ads had the best conversion rate this week"
⢠or "where am i wasting spend"
and get real answers from live campaign data without ever opening the meta ads dashboard
and the part that ties it all together and closes the loop:
he set up a memory system that logs every hypothesis and experiment result across ad iterations.
so when he goes back to step one and generates the next batch of variations...
claude automatically pulls in what worked and what didn't from all previous rounds.
the system literally gets smarter every cycle.
that kind of systematic experimentation across hundreds of ads would normally need a dedicated analytics person just to track
the numbers from the doc:
ad creation went from 2 hours to 15 minutes. 10x more creative output.
and he's now testing more variations across more channels than most full marketing teams
a $380 billion company.
and their entire growth marketing operation (not GTM) = just one person and claude code lol
truly unbelievable
Vladimir Pikalov retweeted
I'm Boris and I created Claude Code. I wanted to quickly share a few tips for using Claude Code, sourced directly from the Claude Code team. The way the team uses Claude is different than how I use it. Remember: there is no one right way to use Claude Code -- everyones' setup is different. You should experiment to see what works for you!
Vladimir Pikalov retweeted
A few random notes from claude coding quite a bit last few weeks.
Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent.
IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits.
Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased.
Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion.
Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage.
Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building.
Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it.
Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements.
Questions. A few of the questions on my mind:
- What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*.
- Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro).
- What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music?
- How much of society is bottlenecked by digital knowledge work?
TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.
Vladimir Pikalov retweeted
š”Tip Thursday - Did you know that the Rill leaderboard is a leaderboard for *every* dimension in the model?š”
The leaderboard is a very powerful part of the Rill interface. Itās a combination of a bar chart + a table. Plus if you have the period comparison selected, you can easily see changes by value and percent.
Check out all the tutorials on YouTubeš
youtube.com/watch?v=aQQBFHbLā¦
Vladimir Pikalov retweeted
"BI-as-code is becoming real.
Last week, I walked through a live demo of one of these tools where:
⢠Metrics layer, semantic model, etc, are both UI & YAML
⢠Everything is version-controlled
⢠GUI & YML = friendly interfaces for everyone
Combine it with a modern analytics database, and youāve got a high-speed, low-friction alternative to legacy dashboardsāideal for financial data, time series, and citizen developer workflows.
This isnāt just faster. Itās reproducible, portable, and explainable."
- John Wessel, Co-Host of the @DataStackShow
Vladimir Pikalov retweeted
Tired: elaborate docs pages for your product/service/library with fancy color palettes, branding, animations, transitions, dark mode, ā¦
Wired: one single docs .md file and a ācopy to clipboardā button.
Vladimir Pikalov retweeted
Today I was sent the following cool demo:
Two AI agents on a phone call realize theyāre both AI and switch to a superior audio signal ggwave
Vladimir Pikalov retweeted
We just published a near-term development roadmap for the model context protocol (MCP)
Some highlights:
- Remote support (and auth!)
- Reference implementations
- Better package management
- Agent support
Vladimir Pikalov retweeted
šStructured Report Generation Blueprint with NVIDIA AI (Llama 3.3)
In this NVIDIA blueprint, we show how to build an agent that can orchestrate the end-to-end process of report planning, web research, and writing. We show that this agent can produce reports of varying and easily configurable format.
We build this agent using Llama 3.3-70b, the most recent of Meta's Llama models that matches the performance of 3.2-406b with 6x fewer parameters. We use NVIDIA NIM's inference service to access this model, taking advantage of LangChain integration with NIM.
Blog: blog.langchain.dev/structureā¦
YouTube: youtu.be/E04rFNtwFcA
Vladimir Pikalov retweeted
Few Shot Prompting š¤ LangSmith Datasets
LangSmith now supports fast, easy-to-use APIs to find few shot examples from your datasets, which you can use to improve your LLM applications without finicky prompt engineering.
You can index dataset examples in one click and dynamically select the most relevant few-shot examples based on user input, optimizing performance for complex apps.
ā”ļø Sign up for the closed beta: bit.ly/langsmith-datasets-feā¦
š Read more in the blog post: blog.langchain.dev/dynamic-fā¦
Vladimir Pikalov retweeted
I wrote last week about why working on a concrete startup or project idea ā meaning a specific product envisioned in enough detail that we can build it for a specific target user ā lets you go faster. In this letter, Iād like to share some best practices for identifying promising ideas.
AI Fund, which I lead, works with many corporate partners to identify ideas, often involving applications of AI to the companyās domain. Because AI is applicable to numerous sectors such as retail, energy, logistics and finance, Iāve found working with domain experts who know these areas well immensely helpful for identifying what applications are worth building in these areas.
Our brainstorming process starts with recommending that a large number of key contributors at our partner corporation (at least 10 but sometimes well over 100) gain a non-technical, business-level understanding of AI and what it can and canāt do. Taking DeepLearning.AIās āGenerative AI for Everyoneā course is a popular option, after which a company is well positioned to assign a small team to coordinate a brainstorming process, followed by a prioritization exercise to pick what to work on. The brainstorming process can be supported by a task-based analysis of jobs in which we decompose employeesā jobs into tasks to identify which ones might be automated or augmented using AI.
Here are some best practices for these activities:
(i) Trust the domain expertās gut. A domain expert who has worked for years in a particular sector will have well honed instincts that let them make leaps that would take a non-expert weeks of research.
Letās say weāre working with a financial services expert and have developed a vague idea (ābuild a chatbot for financial adviceā). To turn this into a concrete idea, we might need to answer questions such as what areas of finance to target (should we focus on budgeting, investing, or insurance?) and what types of user to serve (fresh graduates, mortgage applicants, new parents, or retirees?) Even a domain expert who has spent years giving financial advice might not know the best answer, but a choice made via their gut gives a quick way to get to one plausible concrete idea. Of course, if market-research data can be obtained quickly to support this decision, we should take advantage of it. But to avoid slowing down too much, weāve found that expertsā gut reactions work well and are a quick way to make decisions.
So, if Iām handed a non-concrete idea, I often ask a domain expert to use their gut ā and nothing else ā to quickly make decisions as needed to make the idea concrete. The resulting idea is only a starting point to be tweaked over time. If, in the discussion, the domain expert picks one option but seems very hesitant to disregard a different option, then we can also keep the second option as a back-up that we can quickly pivot to if the initial one no longer looks promising.
(ii) Generate many ideas. I usually suggest coming up with at least 10 ideas; some will come up with over 100, which is even better. The usual brainstorming advice to go for volume rather than quality applies here. Having many ideas is particularly important when it comes to prioritization. If only one idea is seriously considered ā sometimes this happens if a senior executive has an idea they really like and puts this forward as the āmainā idea to be worked on ā thereās a lot of pressure to make this idea work. Even if further investigation discovers problems with it ā for example, market demand turns out to be weak or the technology is very expensive to build ā the team will want to keep trying to make it work so we donāt end up with nothing.
In contrast, when a company has many ideas to choose from, if one starts to look less interesting, itās easy to shift attention to a different one. When many ideas are considered, itās easier to compare them to pick the superior ones. As explained in the book Ideaflow, teams that generate more ideas for evaluation and prioritization end up with better solutions.
Because of this, Iāve found it helpful to run a broad brainstorming process that involves many employees. Specifically, large companies have many people who collectively have a lot of wisdom regarding the business. Having a small core team coordinate the gathering of ideas from a large number of people lets us tap into this collective fountain of invention. Many times Iāve seen a broad effort (involving, say, ~100 people who are knowledgeable about the domain and have a basic understanding of AI) end up with better ideas than a narrow one (involving, say, a handful of top executives).
(iii) Make the evaluation criteria explicit. When evaluating and prioritizing, clear criteria for scoring and ranking ideas helps the team to judge ideas more consistently. Business value and technical feasibility are almost always included. Additionally, many companies will prioritize projects that can be a quick win (to build momentum for their overall AI efforts) or support certain strategic priorities such as growth in a particular part of the business. Making such criteria explicit can help during the idea-generation phase, and itās critical when you evaluate and prioritize.
In large companies, it can take a few weeks to go through a process to gather and prioritize ideas, but this pays off well in identifying valuable, concrete ideas to pursue. AI isnāt useful unless we find appropriate ways to apply it, and I hope these best practices will help you to generate great AI application ideas to work on.
[Original text: deeplearning.ai/the-batch/is⦠]
Vladimir Pikalov retweeted
Guaranteed valid tool calls from OpenAI
Tool calling is incredibly useful for building LLM-steered applications. One of the key difficulties with tool calling is getting models to reliably generate arguments that respect the tool schema. Earlier today OpenAI announced that you can now explicitly enforce that tool call arguments *exactly* match the provided tool schema.
You can try out this functionality in the latest LangChain PY and JS rc releases!
PY docs: python.langchain.com/v0.2/doā¦
JS docs: js.langchain.com/v0.2/docs/iā¦
OpenAI announcement: openai.com/index/introducingā¦
Up next: adding support for the newly announced json_schema response format!
Vladimir Pikalov retweeted
Are you wondering how to build a career in AI?
Then weāve got the perfect eBook for you! It covers necessary steps, including: Developing technical skills, finding projects, nailing the big interview, and more.
Get it here for free: hubs.la/Q02FQBGk0
Vladimir Pikalov retweeted
[LINKEDIN LIVE] Learn how modern iPaaS + flexible API management can consolidate your workflows!
Our Trayers will share how to integrate #AI into your business processes.
Sign up š bit.ly/3x3wjrh #automation
Vladimir Pikalov retweeted
A lot of enterprise developers are building GPTs-like platforms for internal users - let internal users customize the agent to their use case through a UI.
RAGApp is the most comprehensive open-source project available to spin up a RAG/agent chatbot, hosted on any infrastructure you want, without writing a single line of application code - ideal for end users!
In three UI clicks in the admin panel, generate a chatbot over your data with full streaming support, sources, and file attachments (thanks to @nextjs AI SDK)
If youāre a developer, you can of course customize the core ābuilder interfaceā to suit your needs - for instance if you are deploying a platform for your sales associates to use, add custom tools into your CRM, email, custom system prompt templates, and more. Then let the sales associate build the chatbot to their needs.
Check out the RAGApp repo: github.com/ragapp/ragapp
Introducing RAGApp š«
A no-code interface to configure a RAG chatbot, as dead-simple as GPTs by @OpenAI.
Itās a docker container thatās easily deployable in any cloud infrastructure. Best of all, itās fully open-source š„
1ļøā£ Setup the LLM: Configure the model provider (OpenAI, Gemini)
2ļøā£ Setup the data: Define the system prompt and upload your knowledge base.
3ļøā£ Launch the chatbot both via the UI or API
4ļøā£ If via the UI, stream intermediate events and also sources!
This is fantastic work by @MarcusSchiesser and is built upon the same DNA as our create-llama project.
Check out RAGApp today: github.com/ragapp/ragapp
Vladimir Pikalov retweeted
Anthropic has just revealed some exciting news about Claude Sonnet. They've successfully identified how millions of concepts are represented inside this massive model!
LLMs are usually treated as black boxes.
We feed them data, they give us responses, but understanding the 'why' behind their actions is challenging.
Their research used dictionary learning, a technique to find patterns of activated neurons or features.
These features provide a window into the model's understanding of various concepts.
Anthropic mapped out the internal state of Sonnet, finding features for specific entities like cities, people, and atomic elements.
These features are multimodal and multilingual, responding to text and images in multiple languages.
They even found abstract features related to concepts like inner conflict, relationship breakups, and logical inconsistencies.
By manipulating these features, researchers could alter Claude's behavior. For example, activating the "Golden Gate Bridge" feature made Claude obsessed with the bridge, mentioning it in almost every response.
Read more here -> anthropic.com/research/mappiā¦
Vladimir Pikalov retweeted
The best way to learn about Generative AI is by building a project. Weāre excited to introduce a new resource to help you get started!
Our guide to create a Retrieval Augmented Generation (RAG) application provides a detailed project description, a step-by-step outline of the RAG process, and a curated list of @DeepLearningAI short courses that offer both theoretical insights and practical example code.
Check out our guide and start your RAG project: deeplearning.ai/resources/geā¦
Vladimir Pikalov retweeted
Excited to share the latest early release of the early release of our book - Hands-On Generative AI with Transformers and Diffusion Models š„
This release includes new chapters on ML for Audio, fine-tuning LLMs, and fine-tuning SD. Enjoy!
learning.oreilly.com/libraryā¦
Vladimir Pikalov retweeted
You can now generate production-ready prompts in the Anthropic Console.
Describe what you want to achieve, and Claude will use prompt engineering techniques like chain-of-thought reasoning to create more effective, precise and reliable prompts.
Vladimir Pikalov retweeted
šØIntroducing marker-api šØgithub.com/adithya-s-k/markeā¦
Easily convert PDF to MD through an API.
Discovered Marker yesterday by @VikParuchuri, used it, and was hooked. However, I wanted to have an API interface for my project, so I built a wrapper around it.


























