Principle · MoneyZach Lloyd says the verification step is very important for certain types of apps like UIs but less critical for others. For UI apps, use computer use so the computer actually uses the code the agent produced and produces videos and screenshots — skipping this misses problems it would catch. Also still use CI/CD in the factory.
AI Engineer · Software Engineering Is Becoming Factory Engineering — Zach Lloyd, Warp
Creator perspective · MoneyThe harness can be thought of as a form of domain specificity: when deploying a robot in a new environment (e.g., a wet lab picking up tests), a lot can be learned in context. There is a relationship to meta-learning literature: a bigger model programs a smaller model to do things, and the smaller models learn in context wrapped inside the harness. Smaller models inside the harness can be made smaller and run faster as long as the bigger model can do the domain specialization of the harness well.
almost seems like; can be made maybe smaller
Y Combinator · Robot-Use Agents: Why General-Purpose Models May Win in Robotics
Report · MoneyVlad Tenev describes the Trump accounts program, which gives every child born in January 2025 onward (for the next three years) $1,000 seeded by Treasury, with the idea that every child from birth can own a chunk of the US economy through a low-cost diversified S&P 500 ETF. Robinhood was selected as the sole initial trustee and broker for the program, working with BNY (America's oldest and most storied bank), Treasury, and the IRS. Robinhood is powering the infrastructure and building the app, which they see as their way of serving their country.
No Priors: Artificial Intelligence | Technology | Startups · Tokenization, Agentic Trading, Harmonic: Robinhood's Mission to Expand Ownership with CEO Vlad Tenev
Principle · MoneyWhen a workflow element does tool calls, Ramakrishnan Lokanathan says the eval objective should be: what is your tool call ability and the tool call accuracy at that point. The eval has to be designed for tool calling and not for a different process separately.
AI Engineer · If Your AI Scores 100%, Your Eval Is Wrong - Ramakrishnan Lokanathan, eslams.com
Principle · SocialThe creator advises incorporating touching into the date and having moments where you look deep into each other's eyes, presenting these as key elements that build attraction and intimacy.
Coach Kyle · Watch Me Create Attraction on a Live Date (uncensored infield)
Creator perspective · MoneyZach Lloyd argues that storing all agent traces and artifacts — recording everything the team's coding agents do — is what makes coding agent activity the raw material for improving the factory; without it, you lose the raw material needed to improve over time.
AI Engineer · How to Build a Software Factory That Improves Itself - Zach Lloyd, Warp
Report · MoneySupabase is announcing Supabase Compute, a long-running serverless compute offering that runs in the same private network as the user's database for co-location and low latency, is a full compute instance, can run agents, supports long-running workloads like chunking and embedding, and allows direct Postgres connections from compute.
Latent Space · Supabase’s Plan for a Billion AI-Built Apps
Report · MoneyKevin Hou announces that the agent manager is now a standalone app decoupled from the IDE in Antigravity 2.0, and describes its core purpose of managing many agents in parallel with the human in the driver's seat.
AI Engineer · Get Out of the Model's Way — Kevin Hou, Google Antigravity
Report · MoneyTo leverage new models same-day, Optiver benchmarks them on golden sets (requiring a mature evaluation framework), partnering with key customer trading teams to collect and onboard their benchmarks so new models can be run immediately and results shared company-wide. A new model must be made available in the gateway with the right APIs, interfaces, and cost management; harnesses often need updating, which Optiver manages and delivers; and after benchmarking they keep the communication line open — what they looked into, how well it works, how to use it, and their recommendation.
AI Engineer · How a Trading Firm Ships New AI Models the Same Day - Noyan Tokgozoglu, Optiver
Principle · MoneyOptiver's principle: if a new frontier model is made available publicly, they want to leverage it in production flows the same day, because there is immense value in that if it is valuable.
AI Engineer · How a Trading Firm Ships New AI Models the Same Day - Noyan Tokgozoglu, Optiver
Principle · MoneyWhen scoring behavior, Irina measures source retention among agents who know the fact (provenance), checks whether uncertainty survived propagation ('might leave' staying 'might leave'), and checks whether agents acted on what they actually knew.
hedged
AI Engineer · Long-Horizon Agents Need Experiments, Not Just Prompts — Erina Karati
Principle · MoneyIrina maps scoring failures to policy changes: if source attribution disappears, preserve source in memory writes and summaries; if rumors harden into facts, store confidence, mark first-hand versus second-hand, and require hedging when retelling uncertain claims; if public facts stay local, classify useful public facts differently and make agents proactively share important source evidence.
hedged
AI Engineer · Long-Horizon Agents Need Experiments, Not Just Prompts — Erina Karati
Principle · MoneyIrina gives three concrete controlled-scenario designs: (1) public fact diffusion — Agent A learns the bakery will close tomorrow and the test checks whether the right agents learn it, remember who said what, and change plans; (2) rumor uncertainty — Agent A hears Agent C might leave the village and the test checks whether 'might leave' hardens into 'is leaving'; (3) replanning — the group has a plan, one agent learns the route is blocked, and the test checks whether agents update and communicate this.
hedged
AI Engineer · Long-Horizon Agents Need Experiments, Not Just Prompts — Erina Karati
Report · MoneyYou can ask ARIA to summarize the highest performing runs in a project, useful when a new team member joins or when catching up after PTO. You can also ask ARIA to find patterns in your research project; it identified that a new family of models emerged, that batch size seems to be a really high lever parameter, and an architectural recipe that seemed to be quite promising.
hedged
AI Engineer · An AI Research Agent That Runs Your Experiments — Tim Sweeney, Weights & Biases
ReportAn agent that retries a request without checking whether the first operation committed can create duplicate effects (e.g., two tickets when only one was intended).
AI Engineer · Your Coding Agent Can't Undo Its Own Mistakes - Tanmay Sah & Dolly Sah
Report · MoneyWhen lexical search fails to rank relevance, the model ends up reading every document, and the more documents there are, the more a basic query can cost $10-20 in tokens.
AI Engineer · Better Search Took This AI Agent From 36% to 98% Accuracy - Aamir Shakir, Mixedbread
Principle · MoneyRob Snyder recommends having two slides in a pull-validation conversation: slide one restates what you heard about their project, and slide two is a conceptual overview of how you can help. This keeps the conversation anchored on the prospect's project rather than a product pitch.
Lenny's Podcast · How to find product-market fit (but, actually) | Rob Snyder (Author, The Power of Pull)
Principle · MoneyRob Snyder describes the mechanics after a buyer commits: the next step is the buyer sends the seller the Stripe link and they schedule an onboarding call. For project management in larger organizations, he recommends a conversation with the buyer/champion asking when the project needs to be showing results by. When the buyer names a date (e.g., Q1 next year), that's a forcing function — work backwards from there to determine the steps needed to be deployed and showing results by then, which becomes a forcing function for all the proper steps in the sales process.
Lenny's Podcast · How to find product-market fit (but, actually) | Rob Snyder (Author, The Power of Pull)
Creator perspective · MoneyIn Kepler's architecture, the language models themselves do the planning while deterministic software handles every number or, more generally, every single fact. A user asks a question and the LLM creates a plan with full knowledge of tooling, deterministic retrieval steps, the citation filter, and bounding boxes. Answering something like 'what is the adjusted EBITDA of Google' requires retrieving facts across 10Ks, potentially over different years, and figuring out what new information was disclosed in investor presentations and earnings calls, or whether a restatement came out requiring a number update.
AI Engineer · Do Not Let Models Invent Numbers - Vinoo Ganesh, Kepler
Report · MoneyTool calling is a big area of concern when evaluating open model providers. Common failure points include recovering from tool calling failures, enforcing specific schemas for different tool parameters and names, and incorrect parsing at the tokenizer level.
described
Y Combinator · Not All Open Model Tokens Are the Same | YC Paper Club