Files
zen-blog/blog/future-of-open-ai.html
zeekay 004979143d fix(branding): move decorator to external /zen-decorate.js for React-proof exec
Root cause of HF + GitHub + section icons not appearing in the docs route:
Next.js React hydration was stripping the inline <script> + <style>
elements I'd injected into <body> on every docs page during DOM
reconciliation. The script never registered its setInterval.

Move the decorator to /zen-decorate.js loaded via <script defer src>
in <head>. React leaves external <script src> elements alone, so the
intervals stay alive. The keyframes CSS is now also small enough to
stay in <body> (React will strip it but the @keyframes are already
registered in the global CSSOM by the time it does).
2026-05-19 20:08:58 -07:00

181 lines
58 KiB
HTML
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
<!DOCTYPE html><!--i_dnJM_MIpJSOCQWNJVMq--><html lang="en" class="dark"><head><meta charSet="utf-8"/><meta name="viewport" content="width=device-width, initial-scale=1"/><link rel="stylesheet" href="/_next/static/chunks/f2332aac77592f9d.css" data-precedence="next"/><link rel="preload" as="script" fetchPriority="low" href="/_next/static/chunks/e83606e8fa9cc796.js"/><script src="/_next/static/chunks/36d6595f0156cd7e.js" async=""></script><script src="/_next/static/chunks/040e9cea20a8d9c7.js" async=""></script><script src="/_next/static/chunks/d4dffb5a0973d49c.js" async=""></script><script src="/_next/static/chunks/turbopack-967f27a9fd33556d.js" async=""></script><script src="/_next/static/chunks/59d0ad1b64f8544e.js" async=""></script><script src="/_next/static/chunks/4d80e004cf4896dd.js" async=""></script><script src="/_next/static/chunks/350ee4303b732916.js" async=""></script><script src="/_next/static/chunks/8de849ca74fc071f.js" async=""></script><script src="/_next/static/chunks/e62b91212ee7f8ff.js" async=""></script><script src="/_next/static/chunks/f19fe44237e54646.js" async=""></script><script src="/_next/static/chunks/cb0a883bafeb6805.js" async=""></script><meta name="theme-color" media="(prefers-color-scheme: dark)" content="#0A0A0A"/><meta name="theme-color" media="(prefers-color-scheme: light)" content="#fff"/><title>The Future of Open AI — Zen LM Blog | Zen LM</title><meta name="description" content="Reflections on where open AI development is heading and what it will take to get there."/><meta property="og:title" content="Zen LM - Open Foundation Models"/><meta property="og:description" content="Zen AI model family by Hanzo AI. 14 frontier models from 4B to 1T+ parameters for code, reasoning, vision, and multimodal tasks."/><meta property="og:url" content="https://zenlm.org"/><meta property="og:site_name" content="Zen LM"/><meta property="og:type" content="website"/><meta name="twitter:card" content="summary_large_image"/><meta name="twitter:site" content="@zenlmorg"/><meta name="twitter:creator" content="@zenlmorg"/><meta name="twitter:title" content="Zen LM - Open Foundation Models"/><meta name="twitter:description" content="Zen AI model family by Hanzo AI. 14 frontier models from 4B to 1T+ parameters for code, reasoning, vision, and multimodal tasks."/><script src="/_next/static/chunks/a6dad97d9634a72d.js" noModule=""></script><script defer src="/zen-decorate.js"></script></head><body class="antialiased"><div hidden=""><!--$--><!--/$--></div><script>((a,b,c,d,e,f,g,h)=>{let i=document.documentElement,j=["light","dark"];function k(b){var c;(Array.isArray(a)?a:[a]).forEach(a=>{let c="class"===a,d=c&&f?e.map(a=>f[a]||a):e;c?(i.classList.remove(...d),i.classList.add(f&&f[b]?f[b]:b)):i.setAttribute(a,b)}),c=b,h&&j.includes(c)&&(i.style.colorScheme=c)}if(d)k(d);else try{let a=localStorage.getItem(b)||c,d=g&&"system"===a?window.matchMedia("(prefers-color-scheme: dark)").matches?"dark":"light":a;k(d)}catch(a){}})("class","theme","system",null,["light","dark"],null,true,true)</script><!--$--><div data-closed="" role="presentation" hidden="" style="user-select:none;-webkit-user-select:none" class="fixed inset-0 z-50 backdrop-blur-xs bg-fd-overlay data-open:animate-fd-fade-in data-closed:animate-fd-fade-out"></div><div class="bg-fd-secondary/50 p-3 empty:hidden"></div><!--/$--><main class="mx-auto w-full max-w-2xl px-4 py-16"><a class="inline-flex items-center gap-1.5 text-sm text-fd-muted-foreground hover:text-fd-foreground mb-10 transition-colors" href="/blog">← Back to Blog</a><div class="mb-8"><time class="text-xs font-mono text-fd-muted-foreground uppercase tracking-wider">November 10, 2024</time><h1 class="text-3xl font-bold mt-2 mb-3">The Future of Open AI</h1><p class="text-fd-muted-foreground text-lg mb-4">Reflections on where open AI development is heading and what it will take to get there.</p><div class="flex items-center gap-3 pt-4 border-t border-fd-border"><span class="text-sm text-fd-muted-foreground">By <!-- -->Zach Kelling</span><div class="flex gap-1.5 ml-auto"><span class="text-[10px] font-mono uppercase tracking-wider bg-fd-muted text-fd-muted-foreground px-1.5 py-0.5 rounded">Vision</span><span class="text-[10px] font-mono uppercase tracking-wider bg-fd-muted text-fd-muted-foreground px-1.5 py-0.5 rounded">Open Source</span></div></div></div><div class="prose dark:prose-invert max-w-none"><p>Four years into Zen’s development, it’s worth stepping back to assess where we are and where we’re going. Open AI development has made remarkable progress. It also faces significant challenges. Here’s my honest assessment.</p>
<h2 class="flex scroll-m-28 flex-row items-center gap-2" id="what-weve-achieved"><a data-card="" href="#what-weve-achieved" class="peer">What We’ve Achieved</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h2>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="competitive-models"><a data-card="" href="#competitive-models" class="peer">Competitive Models</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Open models now match or exceed proprietary alternatives in many domains:</p>
<ul>
<li><strong>Coding</strong> : Open models lead on HumanEval and MBPP</li>
<li><strong>Reasoning</strong> : Competitive on GSM8K and MATH</li>
<li><strong>General knowledge</strong> : Within 5% on MMLU</li>
<li><strong>Multilingual</strong> : Often superior for non-English languages</li>
</ul>
<p>The capability gap that seemed insurmountable in 2021 has largely closed for models under 100B parameters.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="real-adoption"><a data-card="" href="#real-adoption" class="peer">Real Adoption</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Open models power real applications:</p>
<ul>
<li>Millions of API calls daily through community inference providers</li>
<li>Thousands of fine-tuned variants for specialized tasks</li>
<li>Integration into major development tools</li>
<li>Deployment in production systems worldwide</li>
</ul>
<p>This isn’t just research anymore. It’s infrastructure.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="community-scale"><a data-card="" href="#community-scale" class="peer">Community Scale</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>The open AI community has grown enormously:</p>
<ul>
<li>Hundreds of contributors to major projects</li>
<li>Thousands of researchers building on open foundations</li>
<li>Millions of users accessing open models</li>
<li>Billions of inference requests monthly</li>
</ul>
<p>Network effects are real. Each contribution makes the ecosystem more valuable.</p>
<h2 class="flex scroll-m-28 flex-row items-center gap-2" id="persistent-challenges"><a data-card="" href="#persistent-challenges" class="peer">Persistent Challenges</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h2>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="compute-concentration"><a data-card="" href="#compute-concentration" class="peer">Compute Concentration</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Training frontier models requires massive compute. This compute remains concentrated:</p>
<ul>
<li>Top 5 cloud providers control 80%+ of AI-grade compute</li>
<li>Chip manufacturing bottlenecks persist</li>
<li>Training costs continue rising with scale</li>
</ul>
<p>The Zoo Compute Network helps but isn’t yet large enough for frontier training. We need 10x current capacity.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="data-moats"><a data-card="" href="#data-moats" class="peer">Data Moats</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Large organizations have accumulated unique datasets:</p>
<ul>
<li>Proprietary user interactions</li>
<li>Licensed content</li>
<li>Internal documents</li>
<li>Real-world feedback signals</li>
</ul>
<p>Synthetic data and creative commons content help but don’t fully substitute. Data remains a key differentiator.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="coordination-costs"><a data-card="" href="#coordination-costs" class="peer">Coordination Costs</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Open development is harder to coordinate than corporate efforts:</p>
<ul>
<li>Decisions require consensus</li>
<li>Resources are distributed</li>
<li>Timelines are uncertain</li>
<li>Accountability is diffuse</li>
</ul>
<p>DAOs and governance frameworks help but add overhead. Closed organizations can simply decide and execute.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="safety-complexity"><a data-card="" href="#safety-complexity" class="peer">Safety Complexity</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>As models become more capable, safety becomes more critical:</p>
<ul>
<li>Evaluation is hard (what are we even measuring?)</li>
<li>Mitigation often reduces capability</li>
<li>Adversarial robustness remains elusive</li>
<li>Dual-use concerns complicate openness</li>
</ul>
<p>We can’t ignore safety. We also can’t let it become a barrier to open development.</p>
<h2 class="flex scroll-m-28 flex-row items-center gap-2" id="the-path-forward"><a data-card="" href="#the-path-forward" class="peer">The Path Forward</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h2>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="frontier-open-models"><a data-card="" href="#frontier-open-models" class="peer">Frontier Open Models</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Training truly frontier models openly requires:</p>
<ol>
<li><strong>Scaled compute network</strong> : 100K+ GPU network with reliable coordination</li>
<li><strong>Novel architectures</strong> : Efficiency improvements that reduce compute needs</li>
<li><strong>Better data</strong> : Synthetic data, simulation, and permissioned partnerships</li>
<li><strong>Sustained funding</strong> : Multi-year commitments to training runs</li>
</ol>
<p>This is achievable. It requires coordination the community hasn’t yet demonstrated.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="infrastructure-maturity"><a data-card="" href="#infrastructure-maturity" class="peer">Infrastructure Maturity</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Open infrastructure needs to mature:</p>
<ul>
<li><strong>Training</strong> : Training Gym is solid but needs scaling features</li>
<li><strong>Inference</strong> : Efficient serving remains fragmented</li>
<li><strong>Fine-tuning</strong> : Needs better tooling for non-experts</li>
<li><strong>Evaluation</strong> : Benchmarks don’t capture real-world performance</li>
</ul>
<p>Each layer needs investment. Each layer enables the next.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="governance-evolution"><a data-card="" href="#governance-evolution" class="peer">Governance Evolution</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Current governance works for small decisions. Bigger decisions need:</p>
<ul>
<li><strong>Faster processes</strong> : Weeks, not months</li>
<li><strong>Expert input</strong> : Technical decisions need technical judgment</li>
<li><strong>Accountability</strong> : Approved decisions must be implemented</li>
<li><strong>Representation</strong> : Diverse stakeholders, not just token holders</li>
</ul>
<p>ZIPs are a start. We need to evolve.</p>
<h3 class="flex scroll-m-28 flex-row items-center gap-2" id="safety-by-design"><a data-card="" href="#safety-by-design" class="peer">Safety by Design</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h3>
<p>Open development can lead on safety:</p>
<ul>
<li><strong>Transparency</strong> : Open models can be audited</li>
<li><strong>Diversity</strong> : Multiple approaches reduce correlated failures</li>
<li><strong>Iteration</strong> : Faster feedback loops on safety interventions</li>
<li><strong>Community</strong> : More eyes on potential problems</li>
</ul>
<p>We should demonstrate that openness enhances safety, not undermines it.</p>
<h2 class="flex scroll-m-28 flex-row items-center gap-2" id="predictions-for-2025"><a data-card="" href="#predictions-for-2025" class="peer">Predictions for 2025</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h2>
<p>Concrete predictions to be evaluated:</p>
<ol>
<li><strong>Open models will match GPT-4 class on major benchmarks</strong> (&gt;50% confidence)</li>
<li><strong>Zoo Compute Network will reach 50K GPUs</strong> (60% confidence)</li>
<li><strong>At least one open model will be in top-3 on major leaderboards</strong> (70% confidence)</li>
<li><strong>ZIPs will process &gt;100 proposals</strong> (80% confidence)</li>
<li><strong>Foundation funding will exceed $50M</strong> (40% confidence)</li>
</ol>
<p>I’ll revisit these next November.</p>
<h2 class="flex scroll-m-28 flex-row items-center gap-2" id="a-personal-note"><a data-card="" href="#a-personal-note" class="peer">A Personal Note</a><svg xmlns="http://www.w3.org/2000/svg" width="24" height="24" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100" aria-hidden="true"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"></path><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"></path></svg></h2>
<p>When we started Zen, open AI development was a niche concern. Most assumed the future belonged to well-resourced corporations. That assumption was wrong.</p>
<p>Open development has proven viable. Community coordination works. Distributed resources aggregate. Transparency accelerates progress.</p>
<p>But viability isn’t victory. The next phase requires:</p>
<ul>
<li>Sustained commitment through multi-year projects</li>
<li>Patience with governance overhead</li>
<li>Investment in unsexy infrastructure</li>
<li>Honesty about limitations and failures</li>
</ul>
<p>We’ve shown open AI development is possible. Now we need to show it’s preferable.</p>
<p>The future of AI is still being written. Let’s write it together.</p>
<hr/>
<p><em>Zach Kelling is a co-founder of Zoo Labs Foundation.</em></p></div></main><!--$--><!--/$--><script src="/_next/static/chunks/e83606e8fa9cc796.js" id="_R_" async=""></script><script>(self.__next_f=self.__next_f||[]).push([0])</script><script>self.__next_f.push([1,"1:\"$Sreact.fragment\"\n2:I[106,[\"/_next/static/chunks/59d0ad1b64f8544e.js\"],\"RootProvider\"]\n3:I[53113,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"default\"]\n4:I[73211,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"default\"]\n5:I[10086,[\"/_next/static/chunks/59d0ad1b64f8544e.js\",\"/_next/static/chunks/8de849ca74fc071f.js\",\"/_next/static/chunks/e62b91212ee7f8ff.js\",\"/_next/static/chunks/f19fe44237e54646.js\",\"/_next/static/chunks/cb0a883bafeb6805.js\"],\"\"]\n7:I[89923,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"OutletBoundary\"]\n8:\"$Sreact.suspense\"\na:I[89923,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"ViewportBoundary\"]\nc:I[89923,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"MetadataBoundary\"]\ne:I[6998,[\"/_next/static/chunks/4d80e004cf4896dd.js\",\"/_next/static/chunks/350ee4303b732916.js\"],\"default\"]\n:HL[\"/_next/static/chunks/f2332aac77592f9d.css\",\"style\"]\n"])</script><script>self.__next_f.push([1,"0:{\"P\":null,\"b\":\"i-dnJM_MIpJSOCQWNJVMq\",\"c\":[\"\",\"blog\",\"future-of-open-ai\"],\"q\":\"\",\"i\":false,\"f\":[[[\"\",{\"children\":[\"blog\",{\"children\":[[\"slug\",\"future-of-open-ai\",\"d\"],{\"children\":[\"__PAGE__\",{}]}]}]},\"$undefined\",\"$undefined\",true],[[\"$\",\"$1\",\"c\",{\"children\":[[[\"$\",\"link\",\"0\",{\"rel\":\"stylesheet\",\"href\":\"/_next/static/chunks/f2332aac77592f9d.css\",\"precedence\":\"next\",\"crossOrigin\":\"$undefined\",\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-0\",{\"src\":\"/_next/static/chunks/59d0ad1b64f8544e.js\",\"async\":true,\"nonce\":\"$undefined\"}]],[\"$\",\"html\",null,{\"lang\":\"en\",\"className\":\"dark\",\"suppressHydrationWarning\":true,\"children\":[\"$\",\"body\",null,{\"className\":\"antialiased\",\"children\":[\"$\",\"$L2\",null,{\"children\":[\"$\",\"$L3\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L4\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":[[\"$\",\"main\",null,{\"className\":\"flex min-h-screen flex-col items-center justify-center px-4 text-center\",\"children\":[[\"$\",\"div\",null,{\"className\":\"mb-8 opacity-20\",\"children\":[\"$\",\"svg\",null,{\"width\":\"120\",\"height\":\"120\",\"viewBox\":\"0 0 120 120\",\"fill\":\"none\",\"aria-hidden\":\"true\",\"children\":[\"$\",\"circle\",null,{\"cx\":\"60\",\"cy\":\"60\",\"r\":\"50\",\"stroke\":\"currentColor\",\"strokeWidth\":\"3\",\"strokeLinecap\":\"round\",\"strokeDasharray\":\"280 40\"}]}]}],[\"$\",\"p\",null,{\"className\":\"text-xs font-mono uppercase tracking-[0.3em] text-fd-muted-foreground mb-4\",\"children\":\"404\"}],[\"$\",\"h1\",null,{\"className\":\"text-3xl font-semibold mb-3\",\"children\":\"Page not found\"}],[\"$\",\"p\",null,{\"className\":\"text-fd-muted-foreground max-w-sm mb-10\",\"children\":\"This page doesn't exist, or it may have moved. Try the documentation or head home.\"}],[\"$\",\"div\",null,{\"className\":\"flex flex-wrap gap-3 justify-center\",\"children\":[[\"$\",\"$L5\",null,{\"href\":\"/\",\"className\":\"inline-flex items-center gap-2 rounded-lg bg-fd-primary text-fd-primary-foreground px-4 py-2 text-sm font-medium hover:opacity-90 transition\",\"children\":\"Go home\"}],[\"$\",\"$L5\",null,{\"href\":\"/docs\",\"className\":\"inline-flex items-center gap-2 rounded-lg border border-fd-border bg-fd-background px-4 py-2 text-sm font-medium hover:bg-fd-muted transition\",\"children\":\"Documentation\"}],[\"$\",\"$L5\",null,{\"href\":\"/docs/models\",\"className\":\"inline-flex items-center gap-2 rounded-lg border border-fd-border bg-fd-background px-4 py-2 text-sm font-medium hover:bg-fd-muted transition\",\"children\":\"Browse models\"}]]}],[\"$\",\"p\",null,{\"className\":\"mt-16 text-xs font-mono text-fd-muted-foreground opacity-50\",\"children\":\"zenlm.org\"}]]}],[]],\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]}]}]}]]}],{\"children\":[[\"$\",\"$1\",\"c\",{\"children\":[null,[\"$\",\"$L3\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L4\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]]}],{\"children\":[[\"$\",\"$1\",\"c\",{\"children\":[null,[\"$\",\"$L3\",null,{\"parallelRouterKey\":\"children\",\"error\":\"$undefined\",\"errorStyles\":\"$undefined\",\"errorScripts\":\"$undefined\",\"template\":[\"$\",\"$L4\",null,{}],\"templateStyles\":\"$undefined\",\"templateScripts\":\"$undefined\",\"notFound\":\"$undefined\",\"forbidden\":\"$undefined\",\"unauthorized\":\"$undefined\"}]]}],{\"children\":[[\"$\",\"$1\",\"c\",{\"children\":[\"$L6\",[[\"$\",\"script\",\"script-0\",{\"src\":\"/_next/static/chunks/8de849ca74fc071f.js\",\"async\":true,\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-1\",{\"src\":\"/_next/static/chunks/e62b91212ee7f8ff.js\",\"async\":true,\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-2\",{\"src\":\"/_next/static/chunks/f19fe44237e54646.js\",\"async\":true,\"nonce\":\"$undefined\"}],[\"$\",\"script\",\"script-3\",{\"src\":\"/_next/static/chunks/cb0a883bafeb6805.js\",\"async\":true,\"nonce\":\"$undefined\"}]],[\"$\",\"$L7\",null,{\"children\":[\"$\",\"$8\",null,{\"name\":\"Next.MetadataOutlet\",\"children\":\"$@9\"}]}]]}],{},null,false,false]},null,false,false]},null,false,false]},null,false,false],[\"$\",\"$1\",\"h\",{\"children\":[null,[\"$\",\"$La\",null,{\"children\":\"$Lb\"}],[\"$\",\"div\",null,{\"hidden\":true,\"children\":[\"$\",\"$Lc\",null,{\"children\":[\"$\",\"$8\",null,{\"name\":\"Next.Metadata\",\"children\":\"$Ld\"}]}]}],null]}],false]],\"m\":\"$undefined\",\"G\":[\"$e\",[]],\"S\":true}\n"])</script><script>self.__next_f.push([1,"6:[\"$\",\"main\",null,{\"className\":\"mx-auto w-full max-w-2xl px-4 py-16\",\"children\":[[\"$\",\"$L5\",null,{\"href\":\"/blog\",\"className\":\"inline-flex items-center gap-1.5 text-sm text-fd-muted-foreground hover:text-fd-foreground mb-10 transition-colors\",\"children\":\"← Back to Blog\"}],[\"$\",\"div\",null,{\"className\":\"mb-8\",\"children\":[[\"$\",\"time\",null,{\"className\":\"text-xs font-mono text-fd-muted-foreground uppercase tracking-wider\",\"children\":\"November 10, 2024\"}],[\"$\",\"h1\",null,{\"className\":\"text-3xl font-bold mt-2 mb-3\",\"children\":\"The Future of Open AI\"}],[\"$\",\"p\",null,{\"className\":\"text-fd-muted-foreground text-lg mb-4\",\"children\":\"Reflections on where open AI development is heading and what it will take to get there.\"}],[\"$\",\"div\",null,{\"className\":\"flex items-center gap-3 pt-4 border-t border-fd-border\",\"children\":[[\"$\",\"span\",null,{\"className\":\"text-sm text-fd-muted-foreground\",\"children\":[\"By \",\"Zach Kelling\"]}],[\"$\",\"div\",null,{\"className\":\"flex gap-1.5 ml-auto\",\"children\":[[\"$\",\"span\",\"Vision\",{\"className\":\"text-[10px] font-mono uppercase tracking-wider bg-fd-muted text-fd-muted-foreground px-1.5 py-0.5 rounded\",\"children\":\"Vision\"}],[\"$\",\"span\",\"Open Source\",{\"className\":\"text-[10px] font-mono uppercase tracking-wider bg-fd-muted text-fd-muted-foreground px-1.5 py-0.5 rounded\",\"children\":\"Open Source\"}]]}]]}]]}],[\"$\",\"div\",null,{\"className\":\"prose dark:prose-invert max-w-none\",\"children\":[[\"$\",\"p\",null,{\"children\":\"Four years into Zen’s development, it’s worth stepping back to assess where we are and where we’re going. Open AI development has made remarkable progress. It also faces significant challenges. Here’s my honest assessment.\"}],\"\\n\",[\"$\",\"h2\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"what-weve-achieved\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#what-weve-achieved\",\"className\":\"peer\",\"children\":\"What We’ve Achieved\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}],\"\\n\",[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"competitive-models\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#competitive-models\",\"className\":\"peer\",\"children\":\"Competitive Models\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}],\"\\n\",[\"$\",\"p\",null,{\"children\":\"Open models now match or exceed proprietary alternatives in many domains:\"}],\"\\n\",[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Coding\"}],\" : Open models lead on HumanEval and MBPP\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Reasoning\"}],\" : Competitive on GSM8K and MATH\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"General knowledge\"}],\" : Within 5% on MMLU\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Multilingual\"}],\" : Often superior for non-English languages\"]}],\"\\n\"]}],\"\\n\",[\"$\",\"p\",null,{\"children\":\"The capability gap that seemed insurmountable in 2021 has largely closed for models under 100B parameters.\"}],\"\\n\",[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"real-adoption\",\"children\":[\"$Lf\",\"$L10\"]}],\"\\n\",\"$L11\",\"\\n\",\"$L12\",\"\\n\",\"$L13\",\"\\n\",\"$L14\",\"\\n\",\"$L15\",\"\\n\",\"$L16\",\"\\n\",\"$L17\",\"\\n\",\"$L18\",\"\\n\",\"$L19\",\"\\n\",\"$L1a\",\"\\n\",\"$L1b\",\"\\n\",\"$L1c\",\"\\n\",\"$L1d\",\"\\n\",\"$L1e\",\"\\n\",\"$L1f\",\"\\n\",\"$L20\",\"\\n\",\"$L21\",\"\\n\",\"$L22\",\"\\n\",\"$L23\",\"\\n\",\"$L24\",\"\\n\",\"$L25\",\"\\n\",\"$L26\",\"\\n\",\"$L27\",\"\\n\",\"$L28\",\"\\n\",\"$L29\",\"\\n\",\"$L2a\",\"\\n\",\"$L2b\",\"\\n\",\"$L2c\",\"\\n\",\"$L2d\",\"\\n\",\"$L2e\",\"\\n\",\"$L2f\",\"\\n\",\"$L30\",\"\\n\",\"$L31\",\"\\n\",\"$L32\",\"\\n\",\"$L33\",\"\\n\",\"$L34\",\"\\n\",\"$L35\",\"\\n\",\"$L36\",\"\\n\",\"$L37\",\"\\n\",\"$L38\",\"\\n\",\"$L39\",\"\\n\",\"$L3a\",\"\\n\",\"$L3b\",\"\\n\",\"$L3c\",\"\\n\",\"$L3d\",\"\\n\",\"$L3e\",\"\\n\",\"$L3f\",\"\\n\",\"$L40\",\"\\n\",\"$L41\",\"\\n\",\"$L42\",\"\\n\",\"$L43\",\"\\n\",\"$L44\",\"\\n\",\"$L45\",\"\\n\",\"$L46\"]}]]}]\n"])</script><script>self.__next_f.push([1,"f:[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#real-adoption\",\"className\":\"peer\",\"children\":\"Real Adoption\"}]\n10:[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]\n11:[\"$\",\"p\",null,{\"children\":\"Open models power real applications:\"}]\n12:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Millions of API calls daily through community inference providers\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Thousands of fine-tuned variants for specialized tasks\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Integration into major development tools\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Deployment in production systems worldwide\"}],\"\\n\"]}]\n13:[\"$\",\"p\",null,{\"children\":\"This isn’t just research anymore. It’s infrastructure.\"}]\n14:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"community-scale\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#community-scale\",\"className\":\"peer\",\"children\":\"Community Scale\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n15:[\"$\",\"p\",null,{\"children\":\"The open AI community has grown enormously:\"}]\n16:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Hundreds of contributors to major projects\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Thousands of researchers building on open foundations\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Millions of users accessing open models\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Billions of inference requests monthly\"}],\"\\n\"]}]\n17:[\"$\",\"p\",null,{\"children\":\"Network effects are real. Each contribution makes the ecosystem more valuable.\"}]\n18:[\"$\",\"h2\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"persistent-challenges\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#persistent-challenges\",\"className\":\"peer\",\"children\":\"Persistent Challenges\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n19:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"compute-concentration\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#compute-concentration\",\"className\":\"peer\",\"children\":\"Compute Concentration\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72"])</script><script>self.__next_f.push([1," 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n1a:[\"$\",\"p\",null,{\"children\":\"Training frontier models requires massive compute. This compute remains concentrated:\"}]\n1b:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Top 5 cloud providers control 80%+ of AI-grade compute\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Chip manufacturing bottlenecks persist\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Training costs continue rising with scale\"}],\"\\n\"]}]\n1c:[\"$\",\"p\",null,{\"children\":\"The Zoo Compute Network helps but isn’t yet large enough for frontier training. We need 10x current capacity.\"}]\n1d:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"data-moats\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#data-moats\",\"className\":\"peer\",\"children\":\"Data Moats\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n1e:[\"$\",\"p\",null,{\"children\":\"Large organizations have accumulated unique datasets:\"}]\n1f:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Proprietary user interactions\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Licensed content\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Internal documents\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Real-world feedback signals\"}],\"\\n\"]}]\n20:[\"$\",\"p\",null,{\"children\":\"Synthetic data and creative commons content help but don’t fully substitute. Data remains a key differentiator.\"}]\n21:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"coordination-costs\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#coordination-costs\",\"className\":\"peer\",\"children\":\"Coordination Costs\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n22:[\"$\",\"p\",null,{\"children\":\"Open development is harder to coordinate than corporate efforts:\"}]\n23:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Decisions require consensus\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Resources are distributed\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Timelines are uncertain\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Accountability is diffuse\"}],\"\\n\"]}]\n24:[\"$\",\"p\",null,{\"children\":\"DAOs and governance frameworks help but add overhead. Closed organizations can simply decide and execute.\"}]\n25:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"safety-complexity\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#safety-complexity\",\"className\":\"peer\",\"children\":\"Safety Complexity\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}"])</script><script>self.__next_f.push([1,"]\n26:[\"$\",\"p\",null,{\"children\":\"As models become more capable, safety becomes more critical:\"}]\n27:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Evaluation is hard (what are we even measuring?)\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Mitigation often reduces capability\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Adversarial robustness remains elusive\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Dual-use concerns complicate openness\"}],\"\\n\"]}]\n28:[\"$\",\"p\",null,{\"children\":\"We can’t ignore safety. We also can’t let it become a barrier to open development.\"}]\n29:[\"$\",\"h2\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"the-path-forward\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#the-path-forward\",\"className\":\"peer\",\"children\":\"The Path Forward\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n2a:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"frontier-open-models\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#frontier-open-models\",\"className\":\"peer\",\"children\":\"Frontier Open Models\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n2b:[\"$\",\"p\",null,{\"children\":\"Training truly frontier models openly requires:\"}]\n2c:[\"$\",\"ol\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Scaled compute network\"}],\" : 100K+ GPU network with reliable coordination\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Novel architectures\"}],\" : Efficiency improvements that reduce compute needs\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Better data\"}],\" : Synthetic data, simulation, and permissioned partnerships\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Sustained funding\"}],\" : Multi-year commitments to training runs\"]}],\"\\n\"]}]\n2d:[\"$\",\"p\",null,{\"children\":\"This is achievable. It requires coordination the community hasn’t yet demonstrated.\"}]\n2e:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"infrastructure-maturity\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#infrastructure-maturity\",\"className\":\"peer\",\"children\":\"Infrastructure Maturity\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n2f:[\"$\",\"p\",null,{\"children\":\"Open infrastructure needs to mature:\"}]\n30:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Training\"}],\" : Training Gym is solid but needs scaling features\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,"])</script><script>self.__next_f.push([1,"{\"children\":\"Inference\"}],\" : Efficient serving remains fragmented\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Fine-tuning\"}],\" : Needs better tooling for non-experts\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Evaluation\"}],\" : Benchmarks don’t capture real-world performance\"]}],\"\\n\"]}]\n31:[\"$\",\"p\",null,{\"children\":\"Each layer needs investment. Each layer enables the next.\"}]\n32:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"governance-evolution\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#governance-evolution\",\"className\":\"peer\",\"children\":\"Governance Evolution\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n33:[\"$\",\"p\",null,{\"children\":\"Current governance works for small decisions. Bigger decisions need:\"}]\n34:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Faster processes\"}],\" : Weeks, not months\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Expert input\"}],\" : Technical decisions need technical judgment\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Accountability\"}],\" : Approved decisions must be implemented\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Representation\"}],\" : Diverse stakeholders, not just token holders\"]}],\"\\n\"]}]\n35:[\"$\",\"p\",null,{\"children\":\"ZIPs are a start. We need to evolve.\"}]\n36:[\"$\",\"h3\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"safety-by-design\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#safety-by-design\",\"className\":\"peer\",\"children\":\"Safety by Design\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n37:[\"$\",\"p\",null,{\"children\":\"Open development can lead on safety:\"}]\n38:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Transparency\"}],\" : Open models can be audited\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Diversity\"}],\" : Multiple approaches reduce correlated failures\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Iteration\"}],\" : Faster feedback loops on safety interventions\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Community\"}],\" : More eyes on potential problems\"]}],\"\\n\"]}]\n39:[\"$\",\"p\",null,{\"children\":\"We should demonstrate that openness enhances safety, not undermines it.\"}]\n3a:[\"$\",\"h2\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"predictions-for-2025\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#predictions-for-2025\",\"className\":\"peer\",\"children\":\"Predictions for 2025\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":["])</script><script>self.__next_f.push([1,"[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n3b:[\"$\",\"p\",null,{\"children\":\"Concrete predictions to be evaluated:\"}]\n3c:[\"$\",\"ol\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Open models will match GPT-4 class on major benchmarks\"}],\" (\u003e50% confidence)\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Zoo Compute Network will reach 50K GPUs\"}],\" (60% confidence)\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"At least one open model will be in top-3 on major leaderboards\"}],\" (70% confidence)\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"ZIPs will process \u003e100 proposals\"}],\" (80% confidence)\"]}],\"\\n\",[\"$\",\"li\",null,{\"children\":[[\"$\",\"strong\",null,{\"children\":\"Foundation funding will exceed $50M\"}],\" (40% confidence)\"]}],\"\\n\"]}]\n3d:[\"$\",\"p\",null,{\"children\":\"I’ll revisit these next November.\"}]\n3e:[\"$\",\"h2\",null,{\"className\":\"flex scroll-m-28 flex-row items-center gap-2\",\"id\":\"a-personal-note\",\"children\":[[\"$\",\"a\",null,{\"data-card\":\"\",\"href\":\"#a-personal-note\",\"className\":\"peer\",\"children\":\"A Personal Note\"}],[\"$\",\"svg\",null,{\"ref\":\"$undefined\",\"xmlns\":\"http://www.w3.org/2000/svg\",\"width\":24,\"height\":24,\"viewBox\":\"0 0 24 24\",\"fill\":\"none\",\"stroke\":\"currentColor\",\"strokeWidth\":2,\"strokeLinecap\":\"round\",\"strokeLinejoin\":\"round\",\"className\":\"lucide lucide-link size-3.5 shrink-0 text-fd-muted-foreground opacity-0 transition-opacity peer-hover:opacity-100\",\"aria-hidden\":true,\"children\":[[\"$\",\"path\",\"1cjeqo\",{\"d\":\"M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71\"}],[\"$\",\"path\",\"19qd67\",{\"d\":\"M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71\"}],\"$undefined\"]}]]}]\n3f:[\"$\",\"p\",null,{\"children\":\"When we started Zen, open AI development was a niche concern. Most assumed the future belonged to well-resourced corporations. That assumption was wrong.\"}]\n40:[\"$\",\"p\",null,{\"children\":\"Open development has proven viable. Community coordination works. Distributed resources aggregate. Transparency accelerates progress.\"}]\n41:[\"$\",\"p\",null,{\"children\":\"But viability isn’t victory. The next phase requires:\"}]\n42:[\"$\",\"ul\",null,{\"children\":[\"\\n\",[\"$\",\"li\",null,{\"children\":\"Sustained commitment through multi-year projects\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Patience with governance overhead\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Investment in unsexy infrastructure\"}],\"\\n\",[\"$\",\"li\",null,{\"children\":\"Honesty about limitations and failures\"}],\"\\n\"]}]\n43:[\"$\",\"p\",null,{\"children\":\"We’ve shown open AI development is possible. Now we need to show it’s preferable.\"}]\n44:[\"$\",\"p\",null,{\"children\":\"The future of AI is still being written. Let’s write it together.\"}]\n45:[\"$\",\"hr\",null,{}]\n46:[\"$\",\"p\",null,{\"children\":[\"$\",\"em\",null,{\"children\":\"Zach Kelling is a co-founder of Zoo Labs Foundation.\"}]}]\n"])</script><script>self.__next_f.push([1,"b:[[\"$\",\"meta\",\"0\",{\"charSet\":\"utf-8\"}],[\"$\",\"meta\",\"1\",{\"name\":\"viewport\",\"content\":\"width=device-width, initial-scale=1\"}],[\"$\",\"meta\",\"2\",{\"name\":\"theme-color\",\"media\":\"(prefers-color-scheme: dark)\",\"content\":\"#0A0A0A\"}],[\"$\",\"meta\",\"3\",{\"name\":\"theme-color\",\"media\":\"(prefers-color-scheme: light)\",\"content\":\"#fff\"}]]\n"])</script><script>self.__next_f.push([1,"9:null\nd:[[\"$\",\"title\",\"0\",{\"children\":\"The Future of Open AI — Zen LM Blog | Zen LM\"}],[\"$\",\"meta\",\"1\",{\"name\":\"description\",\"content\":\"Reflections on where open AI development is heading and what it will take to get there.\"}],[\"$\",\"meta\",\"2\",{\"property\":\"og:title\",\"content\":\"Zen LM - Open Foundation Models\"}],[\"$\",\"meta\",\"3\",{\"property\":\"og:description\",\"content\":\"Zen AI model family by Hanzo AI. 14 frontier models from 4B to 1T+ parameters for code, reasoning, vision, and multimodal tasks.\"}],[\"$\",\"meta\",\"4\",{\"property\":\"og:url\",\"content\":\"https://zenlm.org\"}],[\"$\",\"meta\",\"5\",{\"property\":\"og:site_name\",\"content\":\"Zen LM\"}],[\"$\",\"meta\",\"6\",{\"property\":\"og:type\",\"content\":\"website\"}],[\"$\",\"meta\",\"7\",{\"name\":\"twitter:card\",\"content\":\"summary_large_image\"}],[\"$\",\"meta\",\"8\",{\"name\":\"twitter:site\",\"content\":\"@zenlmorg\"}],[\"$\",\"meta\",\"9\",{\"name\":\"twitter:creator\",\"content\":\"@zenlmorg\"}],[\"$\",\"meta\",\"10\",{\"name\":\"twitter:title\",\"content\":\"Zen LM - Open Foundation Models\"}],[\"$\",\"meta\",\"11\",{\"name\":\"twitter:description\",\"content\":\"Zen AI model family by Hanzo AI. 14 frontier models from 4B to 1T+ parameters for code, reasoning, vision, and multimodal tasks.\"}]]\n"])</script><script>
(function(){
// After React renders the lotus emoji from RSC payload, walk the DOM
// and replace it with the enso SVG. Polls so subsequent React re-renders
// also get cleaned up.
function swap(){
// 1. Replace lotus + ' Zen LM' text inside <a> brand links
document.querySelectorAll('a').forEach(function(a){
if (a.dataset.zenLotusSwapped) return;
var t = (a.textContent || '');
if (!/.*Zen LM|Zen LM.*/.test(t)) return;
// Find the lotus text node specifically
var walker = document.createTreeWalker(a, NodeFilter.SHOW_TEXT, null);
var node;
while ((node = walker.nextNode())) {
if (node.nodeValue && node.nodeValue.indexOf('') !== -1) {
node.nodeValue = node.nodeValue.replace(/\s?/g, '');
}
}
// Inject enso SVG at start
if (!a.querySelector('svg[aria-label="Zen LM"]')) {
var ns = 'http://www.w3.org/2000/svg';
var svg = document.createElementNS(ns, 'svg');
svg.setAttribute('viewBox', '0 0 512 512');
svg.setAttribute('fill', 'none');
svg.setAttribute('width', '18');
svg.setAttribute('height', '18');
svg.setAttribute('aria-label', 'Zen LM');
svg.style.cssText = 'display:inline-block;vertical-align:-2px;margin-right:6px;flex-shrink:0';
var p = document.createElementNS(ns, 'path');
p.setAttribute('d', 'M 256 84 A 172 172 0 1 0 256 428 A 172 172 0 1 0 256 84 Z');
p.setAttribute('stroke', 'currentColor');
p.setAttribute('stroke-width', '26');
p.setAttribute('stroke-linecap', 'round');
p.setAttribute('stroke-linejoin', 'round');
svg.appendChild(p);
a.insertBefore(svg, a.firstChild);
}
a.dataset.zenLotusSwapped = '1';
});
}
function start(){ swap(); setInterval(swap, 600); }
if (document.readyState !== 'loading') start();
else document.addEventListener('DOMContentLoaded', start);
})();
</script>
<style id="zenlm-brand-anim">
@keyframes zenlm-brand-draw { to { stroke-dashoffset: 0; } }
svg.zenlm-brand-svg .zenlm-brand-path {
stroke-dasharray: 1100;
stroke-dashoffset: 1100;
animation: zenlm-brand-draw 1.4s cubic-bezier(.45,0,.2,1) 0.1s 1 forwards;
}
.zen-link-icon { display: inline-block; vertical-align: middle; flex-shrink: 0; }
</style></body></html>