Meta just dropped Meta Muse Spark, its first proprietary AI model from the new Meta Superintelligence Labs division. The model itself is interesting, sure. But the real story is what it tells us about where AI is heading for businesses like yours.
Meta Muse Spark scores 58% on Humanity's Last Exam in "Contemplating mode" with tools (50.2% without), and competes directly with OpenAI's GPT Pro and Google's Gemini Deep Think on reasoning benchmarks. Meta says it's over 10x more efficient than their previous Llama 4 Maverick model.
The part most coverage skips over: Meta is spending between $115 billion and $135 billion on AI this year. Nearly double last year. OpenAI just raised at an $852 billion valuation. Google keeps pouring billions into Gemini. Anthropic keeps shipping new Claude models.
For Malaysian businesses, this spending war is actually great news.
How does the AI arms race affect Malaysian businesses?
When tech giants compete this aggressively, prices drop and features improve. We've already seen it:
- ChatGPT's free tier got significantly more capable as Google ramped up Gemini
- Claude Opus pricing fell 67% between versions as providers undercut each other
- Meta AI is free across WhatsApp, Instagram, and Facebook, which millions of Malaysians use daily
More competition means cheaper AI tools. Malaysian SMEs that couldn't justify enterprise AI subscriptions a year ago can now access near-frontier capabilities at a fraction of the cost. Some of them are free.
This matters especially in Malaysia, where AI is expected to generate US$115 billion in productive capacity as part of the country's push to become a regional AI hub by 2030.
Why Meta Muse Spark going proprietary matters more than the benchmarks
Most people aren't talking about this part. Meta built its AI reputation on open source. Llama was the model that made open AI accessible to everyone. Meta Muse Spark? Proprietary. You can use it on meta.ai and the Meta AI app, but you can't download the weights or run it on your own servers.
This matters for two reasons. First, Meta is now directly competing with OpenAI and Google for paying API customers, which means even more price pressure across the board. Second, if you were building on Llama because it was open and free, you need a backup plan.
Our take: Meta going proprietary isn't surprising. The economics of frontier AI don't work with open source alone. But it's a reminder that vendor lock-in is a real risk in this space. The companies giving you free tools today will charge for them tomorrow.
What should Malaysian businesses do right now?
Don't overthink AI strategy. Treat these tools like utilities.
Use what's free while it's free. Meta Muse Spark already powers Meta AI across WhatsApp, which most Malaysian businesses use daily. Test it for customer queries, content drafts, or internal research. The worst that happens is you learn what works and what doesn't.
Don't lock into one provider. Build workflows that can swap between ChatGPT, Claude, Gemini, and Meta AI. If your team only knows one tool, you lose leverage when pricing changes. We've seen clients save 30-40% on AI integration costs just by comparing providers quarterly.
Watch the API pricing wars. Meta Muse Spark's API is already in private preview, and once it opens up it'll put even more pricing pressure on OpenAI and Google. If you're building software that uses AI (chatbots, document processing, data analysis), the next 12 months will see aggressive price drops. Don't sign long-term contracts. Go month-to-month and renegotiate often.
What Meta Muse Spark tells us about AI's next 12 months
The AI spending war isn't slowing down. Meta's $135 billion bet, OpenAI's $852 billion valuation, and Google's continued Gemini push all point the same direction: AI tools will get dramatically better and cheaper over the next two years.
Malaysian businesses that stay flexible will get the most out of this. The ones that pick a single provider and stop paying attention? They'll end up paying more for less. So test new tools when they come out. Keep your options open. And make sure your team knows more than one platform.
At Gotchaa Lab, we help businesses pick the right AI tools without getting locked in. Talk to us if you want an honest assessment of where AI fits in your operations.
References
- Introducing Muse Spark: Scaling Towards Personal Superintelligence — Meta AI
- Meta debuts new AI model, attempting to catch Google, OpenAI — CNBC
- OpenAI valued at $852 billion after completing $122 billion round — CNBC
- Building AI Leadership for Malaysia's 2030 Intelligent Economy — Asia School of Business
