What happens when AI starts running its own research?
Our CEO @RichardSocher takes the Main Stage at @aiDotEngineer to talk through it.
"First Steps Toward Automated AI Research" July 1, 10:45 AM
See you there.
You.com
3,681 posts
You.com
@youdotcom
The Leading Web Search APIs for AI
Get started → you.com/docs
- What happens when you burn through your entire 2026 AI coding budget in 4 months and your COO then admits he couldn't link token usage to shipped features? That's an architecture problem. 🧵Replying to @youdotcom3 fixes that don't require rearchitecting everything: → Minimal tool schemas (parameters are decision surfaces, they cost tokens) → Payloads that return only what's relevant now → Token budget as an explicit field the agent can readThe metric that matters: cost per task—not total tokens. Tokens per resolved ticket, per merged PR, per clean extraction. Cheaper inference won't fix waste, but measuring the right thing might. Full article:
- @youdotcom search is in @AgnoAgi now. 🔍 Free MCP endpoint—100 queries/day, no signup, no key. Nothing between you and a working web search agent.
- @stripe's new Directory indexes services for AI agents to discover and integrate autonomously—and You.com is in the example on their docs page. 🔎 If you search "free web search" in the Stripe directory, boom, @youdotcom shows up. Is your product findable by
Documentation: docs.stripe.com/directory - 🤖 AI writes the code. So what does a software engineer actually do now? Join @youdotcom, @supersetstudio & @foundersbay Thursday as we unpack this topic live in SF—speakers from @nvidia, Ketch, and a startup studio building AI-native from scratch.
- You.com reposted
Factory helps @youdotcom accelerate their software engineering team by running agents end-to-end across the software development lifecycle, often in the background for days.
- Web search is invisible infrastructure for AI agents—until it fails mid-session. @FactoryAI's Droid agent calls it multiple times per workflow. Their old provider was quietly making every session slower and more expensive. Here's what they found when they benchmarkedReplying to @youdotcomThe quote from @FactoryAI says it plainly: "We cut P50 web search latency from over five seconds to sub-one second and pushed our success rate to three 9s or higher. Sessions cost less because Droid gets to a correct answer faster when it has the right context up front."If your agents depend on the web for grounding or retrieval, the search API underneath matters more than most teams realize—until it doesn't. Full case study →
- You.com reposted
Excited to partner with @youdotcom ! You.com powers search for DuckDuckGo, Amazon, Alibaba, and more. With You.com on Orthogonal, agents can search the web and get LLM-ready results in 300ms, extract clean content from any URL, run multi-step
- Every leap in living standards came from scaling something scarce. We figured out manufacturing. Distribution. Communication. Dedicated intelligence was the last holdout—until now. 🧠
00:00Full video:
