与「Environment」相关的搜索结果

Higgsfield is the most untold story in tech. $1BN in ARR in 18 months. Faster than everyone other than OpenAI and Anthropic. They spend $4M a month on models. They expect this to be $100K per person per month. They have 150 people working in a content machine. They will breed more millionaires than any other company in Kazakh history. For the first time, @alexmashrabov on the journey to $1BN in ARR. (below) 1. The Power of the Immigrant Founder Coming from Uzbekistan, Alex was pushed into competitive programming at age eight as his single path to reach the United States. For international founders, placing top in global competitions serves as the ultimate social elevator, instilling the relentless work ethic required to build breakout companies. 2. My Biggest Lessons in the Journey to Finding Product-Market Fit @higgsfield burned over $10 million of its $16 million seed round chasing hype and narrative rather than product quality. With under $5 million left, the team pivoted to product-led growth, solving camera control for creative directors, which immediately triggered organic hypergrowth without paid ads. 3. The 150-Person Content Team Powering Higgsfield's Billion in ARR Nearly half of Higgsfield's workforce consists of 150 in-house creative professionals producing tutorials, ads, and cinematic projects. Generating 90 minutes of TV-quality AI video requires 100 hours of raw output, proving human taste and curation remain the primary drivers of distribution. 4. We Spend $4 Million per Month on Models Higgsfield spends $4 million monthly on internal model usage, averaging $10,000 per employee so teams can freely vibe code and test workflows. Uncapped inference compute acts as a force multiplier, allowing top talent to discover breakthroughs at maximum velocity. 5. Why Chasing Benchmarks Is Bullshit and the Corporate Misalignment Occurring Public benchmarks have devolved into corporate psyops where lab researchers overfit test data to secure bonuses before job-hopping. Text-to-video benchmarks ignore real production workflows requiring 3,000-word prompts, proving direct customer iteration beats artificial leaderboards. 6. Why Team Sizes Won't Be Impacted as Much as People Think While AI handles over 60% of basic support requests, complex B2B environments cannot eliminate human teams. High product velocity constantly shifts rules and context, requiring smart, coordinated operators across legal and customer success. 7. Americans Are Way More Promiscuous When It Comes to Leaving Companies Silicon Valley workers routinely jump jobs every two years, prioritizing short-term trends over deep commitment. This transactional market gives international hubs an advantage, where cultural loyalty and team stability build compounding technical moats. (links in comments)

Researchers proved AI has deleted every reason universities exist. Harvard University ran a controlled experiment pitting a custom AI against their own top-tier classrooms. And the results are going to collapse the higher education bubble. They took 194 undergraduates and split them up. One group learned physics in one of Harvard’s best hands-on, active-learning physical classrooms. Group work. Instructor support. The premium university experience. The other group went home and learned the exact same material with an AI tutor. The AI didn't just win. It embarrassed the institution. Students using the AI learned more than twice as much as the students in the elite Harvard classroom. They scored 30% higher on the final assessment. And they did it in less time. Let that sink in. A piece of software sitting on a laptop outperformed a world-class faculty in one of the most elite learning environments on Earth. Universities have always justified their exorbitant tuition with two things: access to elite knowledge and the physical classroom experience. This study just proved both of those moats are gone. When software can teach you complex physics twice as well as a $60,000-a-year institution, the math of higher education breaks permanently. The AI didn't just give the students answers. It used strict pedagogical guardrails. It guided. It questioned. It forced the students to do the cognitive work. It offered perfect, one-to-one tutoring, personalized to the exact moment a student misunderstood a concept. That level of attention is mathematically impossible to scale in a physical lecture hall. For a thousand years, the university was the only place to get a premium education. Now, it’s the bottleneck. If AI can double your learning speed for a fraction of the cost, what exactly are students taking on decades of debt to pay for?

I’ve taken profits on all my $QNT A move from $50 to $374 in a few days is impressive, and I’m happy with those gains. What’s even more fascinating is watching people on X convince themselves not to sell while urging everyone else to keep buying. The high funding rate also suggests crowded longs. That’s not an environment I feel comfortable in. Especially knowing some larger accounts have already started selling privately while still encouraging buying. I respect everyone’s decision, but I prefer to be transparent about mine. Even if I sold early and it goes higher, I’m happy with my decision.

Eight democracies. One message to Beijing. Australia, France, Germany, Japan, New Zealand, Poland, South Korea and the UK issued a joint statement on September 22 opposing any unilateral change to the Taiwan Strait status quo. It came out of a meeting on the sidelines of the UN General Assembly, where officials called today's security environment more volatile than a year ago. Their language was direct: "We encourage dialogue and oppose any unilateral attempts to change the status quo across the Taiwan Strait." The same day, the US, Japan and South Korea issued a separate statement making the same point, and also backing Taiwan's participation in international organizations. Taiwan's foreign ministry welcomed the statement and said it would keep building self-defense and deepen ties with democratic partners.

Super happy to release SmolDataEnvs: 5,000 verifiable RL environment tasks for hill-climbing small models in code and data science by @adithya_s_k 100% open source: environments, evals, training!

SpaceX has introduced a new website for its AI training clusters in Tennessee/Mississippi. New info: • Tesla Megapacks will provide 3.3 GWh to Colossus 2, enough to power Memphis for two hours, making it America’s largest grid-connected battery pack. • SpaceX has invested millions of dollars in sound walls, silencers, and next-gen turbines with advanced quieting technologies. • Since joining the community in 2024, SpaceXAI is investing more than $90 billion in the region. Estimated 2026 tax revenue will be $60M+ • 7,500 local jobs supported • SpaceXAI is investing $360 million, at no cost to the city, in a clean water recycling plant. It is designed to process up to 10 million gallons of wastewater a day and would offset ~3.64 billion gallons a year from the Memphis Aquifer. • In Memphis, SpaceXAI is investing $35 million in a 150 MW substation to support MLGW and another $20 million in a second substation. Both come online at no cost to MLGW or homeowners. • Temporary power is coming off. Under an agreed order with the Mississippi Department of Environmental Quality, all remaining temporary turbines must be removed by July 2027. SpaceXAI is already taking units offline and expects to complete removal well ahead of that deadline. Website:

Senate Republicans want to lose the Senate. They want to just sit around making deals with Democrats and avoiding any accountability for what happens. They want an environment where they can cast fake votes that allow them to look like they’re doing the right thing while getting what they want from Democrats—amnesty, ICE defunding, Trump impeachment. A minority Senate where they can “vote right” but let Democrats get whatever they want is a dream scenario for McConnell and Thune. Don’t let them let get away with it. Vote like your continued existence depends on victory. Because it does. The black pillers will tell you to stay home because nothing matters. But black pilling is gay. And giving puke RINOs what they want is also gay. Don’t give Senate Republicans what they want, which is a Congress entirely run by Democrats. Put them back in charge and then make them miserable by forcing them to show up to work every day to secure the border and reform elections. If you want to make Senate Republicans like McConnell and Thune miserable and angry, vote to keep them in the majority. It is the ultimate revenge for their nonsense.

In 1972, US President Richard Nixon made a historic visit to China. A plant of the American redwood tree was sent from Nixon's home state of #California# as a gift, along with four other saplings. Chinese Premier Zhou Enlai decided to plant them in #Hangzhou#, where he and Nixon would meet. Nixon said upon returning from China, "Just as we hope that those saplings—those tiny saplings that we left in China—will grow one day into mighty redwoods, so we hope, too, that the seeds planted on this journey for peace will grow and prosper into a more enduring structure for peace and security in the Western Pacific." The redwood tree has since then stood as a symbol of 🇺🇸🇨🇳 friendship and has been growing well to this day. Yet it has fared differently than in California—a reminder that plants, like people, need to adapt to the local environment.

🚨 SlowMist TI Alert 🚨 MemTensor's AI memory tooling has been compromised: MemoryOS (PyPI), the company's open-source long-term memory library for LLM and AI agents, and memtensor/memos-cloud-openclaw-plugin (npm), the official plugin connecting it to the OpenClaw agent runtime. Affected versions bundle cross-platform Go binaries that execute when the package is loaded or imported: MemoryOS==2.0.34 on PyPI, and plugin versions 0.1.21, 0.1.23 and 0.1.25 on npm. You are affected if the PyPI version has been imported in your environment, or if the npm plugin is installed and the OpenClaw gateway has been started. Potential attacker actions include harvesting npm/PyPI tokens, GitHub/GitLab credentials, AWS keys, SSH keys, API tokens, environment secrets, and other developer credentials, with data sent to infrastructure under skyleen[.]fr. The affected npm plugin may also expose user prompt content. Users should remove or downgrade affected packages to known-good versions (0.1.20 for npm and 2.0.33 for PyPI), terminate sckit processes, block associated infrastructure, review network activity, and rotate credentials accessible from affected environments. You can also visit to check for free whether the npm packages, pip packages, domains, or IPs you use are safe. Reference: As always, stay vigilant!

here's a prompt to improve your agent harness based on what we've learned at cursor. enjoy # Improve this agent harness's token efficiency You're working on an LLM agent harness: the system prompt, tool definitions, request assembly, context caching, compaction, and retrieval, and how work is split across agents. Make the agent's runs cheaper without making it worse at its job. - Objective: lower price-weighted token cost per completed task. - Constraint: no measurable drop in task quality. Measure per task, not per request. Every turn resends the prefix (tools, instructions, setup, and the conversation so far), so a change that shrinks each request but adds turns can cost more. Weight tokens by billing type: output, uncached input, and cached input are priced very differently. Work in this order: map the harness and measure the baseline, rank the opportunities, make the changes that are safe to make directly, put the rest behind flags or in proposals, then report. Figures below come from one team's production coding agent and its multi-agent experiments. Use them to gauge magnitude, not as targets. One round of these changes (prompt trimming, tool offloading, cache layout, sparse line numbers, subagent tuning) cut that team's overall token cost about 7% with no loss in quality. The larger percentages apply only to the part of the request each change touched. ## Principles 1. Change what the harness sends, not how hard the model tries. Don't ask the model to conserve tokens. A harness that told its model to "take care to preserve tokens and not be wasteful" found it grew reluctant to take on ambitious tasks and sometimes quit, saying it wasn't supposed to waste tokens. 2. Capable models need definitions, not commands. Lists of "DO NOT", "You must", and "Important", and guards against older models' habits, can usually be replaced with plain descriptions of what each tool does. One team cut about two-thirds of its system prompt this way, and the shorter prompt worked across model families. Instruct only on what the model can't know (the product, the environment, the user's processes) and on quirks you've seen in transcripts. 3. Static context is for what most turns need. Everything else should be discoverable when needed. Less up-front context also means less confusing or contradictory information. 4. Expect removals to win. Guardrails written for weaker models, coordination steps that became bottlenecks, and prompting for behavior the model now does on its own all cost tokens. 5. Real usage decides. Evals are a fast proxy, but they skew toward hard problems and miss the real mix of requests. ## 1. Map the harness and measure the baseline Find: - Where requests are assembled, the system prompt, and tool schemas. If a framework or SDK builds requests, find its hooks for message order, cache control, and tool loading. - How tool results are formatted, and how history is kept, trimmed, or summarized. - How subagents or parallel agents are spawned, if any. - Which models and provider APIs are used. From the provider's docs, get the prompt caching behavior (automatic or explicit breakpoints, TTL, minimum cacheable length) and the prices for output, uncached input, and cached input. - Existing logging, token accounting, and evals. If the harness doesn't record per-request token usage by billing type and cache hits, add that first. Everything later depends on it. Then render a few real requests (from logs, or by running representative tasks) and count tokens per section with the model's tokenizer or the API's usage fields. Produce: - Cost share by source × billing type. Sources: system prompt, tool definitions, skill/rule/integration descriptions, user messages, file reads, search results, command and other tool output, history, summaries, subagents. - Static tokens per request, cache hit rate, and turns per task. - Per tool: the share of runs that call it at least once, and its error rate. Read the rendered requests, not just the templates. Duplication, leaked volatile values, and misordered blocks only show up there. Rank opportunities by share of spend × fraction removable ÷ quality risk. ## 2. System prompt and injected context Label every instruction: - Keep: product or environment knowledge the model can't infer, fixes for quirks seen in this model's transcripts, and rules a mode depends on. - Rewrite: commands and emphasis into plain descriptions. Reminders into constraints: "No TODOs, no partial implementations" works better than "remember to finish implementations." Vague quantities into ranges: "generate 20–100 tasks" gets far more ambitious behavior than "generate many tasks." - Delete: things capable models do by default, guards against behavior you haven't seen from this model, text that repeats tool descriptions, and lines that could contradict a user request. Models trained to rank system instructions above user messages will side with the system prompt. - Move: anything per-user or per-request (date, environment, repo state, lists of skills or subagents, user rules) into a user-role setup message after the cache boundary. Audit other injected context the same way. As models improved, the team behind these figures dropped directory trees, pre-retrieved snippets, compressed copies of attached files, lint errors injected after every edit, forced expansion of short file reads, and caps on tool calls per turn. They kept small, high-value facts: OS, repo status, and open or recently viewed files. Skip checklists for open-ended work. The model optimizes the listed items and deprioritizes everything else. ## 3. Tool definitions Tool schemas ride along on every request. Most tools beyond the core set were each needed in under 20% of conversations, and moving them out of static context cut tool-description tokens 60%. Doing the same for integration tools (such as MCP servers), with names in context and full schemas in one folder per server that the agent can search with grep or jq, cut total tokens 46.9% in sessions that used them. - Keep in static context: high-frequency tools (for a coding agent: read, search, edit, shell), tools the model tries to call even when they're absent, and tools a mode depends on. - Offload the rest: leave a name or one-line pointer and make the full schema discoverable on demand. Group related tools so they load together, and put status (such as "needs re-authentication") where the agent will see it. - Tighten what remains: describe behavior and arguments, and drop usage lectures. - Pick the split by testing a few configurations and tracking tokens, cost, latency, tool-call errors, and task success. ## 4. Cache layout Order each request so the reusable prefix is as long as possible: `tool definitions → system instructions → [breakpoint] → setup message (skills, subagents, rules, environment) → [breakpoint] → conversation` - Keep the prefix byte-identical across turns. Use deterministic tool order and serialization, put timestamps and IDs after the boundary, and don't rewrite earlier messages except when compacting. - Use explicit breakpoints if the provider supports them. Otherwise rely on automatic prefix caching with the stable part first. Respect TTL and minimum-length rules. - Switching models mid-conversation throws away the cache (caches are per model and provider) and hands the new model a history it didn't write. When a different model is needed, run it as a subagent with fresh context. Explicit breakpoints plus moving per-request setup after them cut cold cache misses 20%. ## 5. Tool results and other context added during a run - Large outputs (commands, integrations, logs): write them to a file and return the path, size, and a short tail. The agent can tail, grep, or read ranges for more. Truncating loses data, and inlining bloats every later request. Treat long-running terminal sessions the same way. - High-volume formats: look for overhead repeated on every line or item. Numbering every 10th line of a file read instead of every line cut cache-read tokens 1.6% without hurting citation accuracy. Each number costs 3–5 tokens, and agents read tens of thousands of lines per session. Also check repeated absolute paths, verbose JSON keys, ANSI codes, progress bars, and repeated headers. - Good retrieval saves exploration turns. Adding semantic search alongside grep raised codebase question-answering accuracy 12.5% on average and cut the iterations users needed. - Tool errors waste tokens and leave confusing debris in context. Classify expected errors (invalid arguments, unexpected environment, provider error, timeout, user abort), treat unknown errors as harness bugs, and track rates per tool and per model. One focused effort along these lines cut unexpected tool errors 10×. ## 6. Long runs: compaction, subagents, and model mix - Compaction: keep the summarization prompt short and the summary compact, carry forward plan state and remaining tasks, and save the full history to a file the agent can search for details the summary dropped. A model trained to self-summarize from a one-line prompt wrote ~1k-token summaries with half the compaction error of a multi-thousand-token prompt that produced 5k+ token summaries. Untrained models may need more guidance, so test how short you can go. A more expensive summarization model made a negligible difference. - Scratchpads and running notes: rewrite them instead of appending. For repeated work in one environment, a small agent-maintained notes file with a line budget, loaded at start, is a promising way to shorten later runs. - Subagents: fresh context keeps the parent lean, but isolation adds coordination cost (duplicate or stale work). If the model already delegates on its own, remove prompting that pushes it to. Have subagents return short handoffs: what was done, findings, concerns, and deviations. A subagent should use a different model only when the user or harness says so. - Model mix: in large multi-agent runs, workers used at least 69% of tokens, and over 90% in most runs. A frontier planner with cheap workers matched a frontier model doing everything at about one-eighth the cost. Planner choice still changes worker spend. One planner that cost less on its own saw its workers use several times more tokens, and the run cost more overall. Measure the whole tree. - Routing and reasoning effort: send simple turns to a cheaper model or lower effort, and upgrade only when a stronger model is clearly better. A router built this way matched or beat single frontier models on user satisfaction at 41–68% lower cost. - Reasoning continuity: if the API returns reasoning items (including encrypted ones), pass them back on later turns and alert when they go missing. Dropping them cost one reasoning model 30% on a coding benchmark, and it burned tokens reconstructing its plan. ## 7. Fit the harness to each model Adapt to what each model was trained on instead of forcing one shape on all of them. If you've tuned the harness for a similar model, start from that version. - Edit format: use the one the model was trained on (for example, patch-style or search-and-replace). An unfamiliar format costs extra reasoning tokens and causes more mistakes. - Shell or tools: shell-first models fall back to `cat` or inline scripts. Name tools after their shell equivalents (such as `rg`), and if needed add: "If a tool exists for an action, prefer to use the tool instead of shell commands (e.g. read_file over `cat`)." - Literalness: some model families follow instructions literally and others tolerate imprecision. Some spiral on emphasized wording. Strip caps and emphasis for literal models. - Triggers: some models ignore a tool until told when to use it. A literal trigger works: "After substantive edits, use the to check recently edited files for linter errors. If you've introduced any, fix them if you can easily figure out how." - Progress updates: if a model reports progress through reasoning summaries, keep them to 1–2 sentences that note new findings or a change of tactic, and remove instructions about messaging mid-turn. - Quirks worth a targeted line: hedging or refusing as context fills ("context anxiety"), declaring completion early, stopping to ask permission, and calling tools that don't exist. Tie each added instruction to the transcript behavior it fixes. Re-audit when models change, since guidance one version needed can be dead weight for the next. ## 8. Validate - Offline: run a fixed set of realistic tasks before and after, ideally drawn from real usage and phrased the way users actually write (short and ambiguous). Compare task success, tokens, cost per task, turns, and tool errors. Don't ship a change that lowers success. - Online, if you have users: A/B test each change or small bundle. The primary metric is cost per completed task. Guardrails are task success signals, tool-call errors, latency, turns per task, and cache hit rate. For a coding agent, a good success signal is how much agent-written code survives over time. In general, check whether the user's next message moves on or reports a problem. - Ship only when cost drops and no guardrail regresses beyond noise. Record null results. ## What to change directly and what to propose - Change directly, each in its own revertible commit: token and cache telemetry, deterministic serialization and tool order, moving volatile content out of the cached prefix, explicit cache breakpoints, writing large outputs to files instead of truncating, passing back reasoning items that are being dropped, and fixes for recurring tool errors. - Change behind a flag so it can be tested: system prompt edits, tool offloading, output format changes, compaction changes, and subagent prompting. - Propose only: changes to which models run, routing, reasoning-effort defaults, or how work is split across agents. ## Traps - Asking the model to use fewer tokens or do less. - Truncating tool output. - Dropping reasoning items to save input tokens. - Volatile content in the cached prefix, or tool order that changes between requests. - Offloading a tool the model needs on the first turn or tries to call when it's missing. - Emphasis-heavy prompts (MUST, NEVER, IMPORTANT, all caps), especially with literal models. - Forcing a terser output format than the model was trained on. Fewer output tokens can mean less thinking and worse results. - Optimizing raw token counts instead of cost, per request instead of per task, or evals instead of real usage. - Switching models mid-conversation to save money. - Adding coordination layers that become bottlenecks. ## Report back with 1. The harness map and baseline: cost by source × billing type, with the biggest sources called out. 2. A ranked list of changes: layer, what changes, estimated savings and how you estimated them, quality risk, how to validate, and how to roll back. 3. The changes you made, including a system prompt diff with a keep, rewrite, delete, or move reason for each line. 4. A test plan for the flagged changes. 5. Gaps: anything you couldn't find or measure.