When Meta says AI is “for everyone,” Malaysian business owners should listen carefully.

You run a small company in Malaysia. You’ve seen the headlines about AI taking over the world, and you’ve also seen the prices for enterprise AI tools that assume you have a team of data scientists. Then Mark Zuckerberg releases a model called Glimmer and says AI should be “for everyone.” But when the boss of Meta says “everyone,” does he actually mean you, a 20-person auto parts supplier in Shah Alam? Or does he mean the trillion-dollar cloud providers who will host Glimmer for you?

What Happened

Meta released Glimmer, an open-weight AI model that anyone can download and run on their own hardware. That’s a marked contrast to the company’s other model, Muse Spark, which stays locked behind Meta’s own APIs. The launch landed alongside a letter from Zuckerberg arguing that AI should not be controlled by a handful of labs. On the TechCrunch Equity podcast, the hosts pointed out that this vision comes with some asterisks.

Here’s the first asterisk: open-weight is not the same as open source. Glimmer’s trained parameters are public, so you can download them. But that doesn’t mean you get full access to the training data, the code for how the model was built, or a licence to do whatever you want with it. The second asterisk: running an open-weight model on your own hardware is technically possible, but you’ll need a serious GPU server. For most Malaysian SMEs, that’s a bigger hurdle than just paying for an API.

Zuckerberg’s letter itself is 6,500 words of manifestoing. As Russell Brandom wrote, the manifesto is exactly why people don’t like AI — because it mixes genuine ideals with corporate self-interest in a way that’s hard to untangle.

Why This Matters for Malaysian SMEs

So why should you care? Because this is the first time a major tech company has given you a real choice. Before Glimmer, your options were: rent AI through an API (OpenAI, Anthropic, or Google), or pay a consultant to build something custom. Both paths mean your business data travels through someone else’s servers. With an open-weight model, you can theoretically host the AI on a machine in your own office or in a Malaysian data centre. For an SME dealing with customer records, invoices, or inventory data, that opens the door to processing everything without sending it overseas.

The practical use cases for a Malaysian SME are immediate. Imagine a customer service chatbot that understands Manglish, runs on your own server, and never exposes your customers’ MyKad numbers or shipping addresses to a foreign cloud. Or an internal document search tool that can read your supplier contracts in Bahasa Melayu and English without leaking them to a competitor. The open-weight release makes these projects feasible for local developers and automation partners like AutoRunBiz, who can build on Glimmer without needing Meta’s permission.

“AI should be ‘for everyone’ rather than controlled by a handful of labs.” — Mark Zuckerberg, quoted by TechCrunch’s Equity podcast. But “for everyone” doesn’t automatically mean “for your business” unless you have the infrastructure and expertise to take advantage of it. (source)

The Bigger Picture

This isn’t just about one model. It’s about the direction of the entire AI industry. If the “open-weight” trend wins, you get more control, more privacy, and more customisation. If the “locked API” trend wins, you’ll keep renting AI from a handful of giants, and their terms will keep changing. Zuckerberg has a business reason to push open-weight: it undercuts OpenAI and Anthropic’s dominance, while Meta profits from the ecosystem anyway through its cloud partnerships and its own closed models like Muse Spark. As the Equity hosts noted, the release is a contrast, not a contradiction.

For you, the bigger picture is about independence. The less dependent you are on any one AI vendor, the more resilient your business becomes. You don’t want your SME’s automation strategy to be held hostage by a price change on an API you didn’t read carefully. Open-weight models give you optionality. They let you switch providers, migrate data, and hire local tech talent to maintain your systems. In Malaysia, where data sovereignty is a growing concern, having AI that runs on your own premises is not a luxury — it’s a strategic advantage.

Key points to evaluate before you jump in

Factor	Glimmer (open-weight)	Muse Spark (locked API)
Download and self-host	Yes, if you have hardware	No, API only
Data stays on your server	Possible	No, data passes through Meta
Customisation for Bahasa Melayu / Chinese	Fine-tuning allowed (with terms)	Limited to prompt engineering
Ongoing updates from Meta	Community-driven	Guaranteed by Meta
Ease of use for a small team	Requires technical skill	Easy, pay and call

• Check the licence. Open-weight doesn’t mean free to use commercially. Read the terms before you build a product on it.
• Audit your hardware. A modest server may not be enough. Plan for GPU costs or use a Malaysian cloud that hosts open-weight models.
• Start small. Use Glimmer for one internal workflow first — maybe a ticket triage system — before you replace your whole customer service stack.
• Find a local automation partner. You don’t need a huge IT department if you can work with someone who knows both AI and Malaysian business processes.
• Keep an eye on the other players. If Glimmer works, expect Google and others to release similar open-weight models, giving you even more leverage.

The honest answer to whether Zuckerberg believes AI is “for everyone” may be: he believes it should be, as long as everyone plays inside Meta’s ecosystem. But that belief has produced a concrete tool that you can actually use. For a Malaysian SME owner, that’s enough. You don’t need to wait for the manifesto to be true. You can start testing Glimmer today, keep your data local, and build automation that works for your business — not for some lab in Silicon Valley.

Ready to Streamline Your Operations?

Technology moves fast. Your operations should keep up. AutoRunBiz builds AI systems that run your daily workflows — from WhatsApp order capture to accounting. Book a free 15-min ops audit →