Should AI be Open-source? Mark Zuckerberg’s Open Letter Explained

By Amanda Kavanagh

In a recent open letter, Meta’s Mark Zuckerberg aired his thoughts about how AI should be open source, and not just known by a handful of big companies. It’s a long-read, so we’ll save you some time by breaking it down.

Zuckerberg begins by drawing parallels between AI and the historical rise of Linux. He highlights that in the early days of high-performance computing, major tech organizations invested in developing their own closed-source versions of Unix.

However, open-source Linux gained popularity as it allowed developers to modify its code, and was more affordable.

As a result, over time Linux became more advanced, and a secure and broad ecosystem developed.

Now it’s the industry standard for cloud computing and the operating systems that run most mobile devices. “And we all benefit from superior products because of it,” he writes.

Zuckerberg's push for open-source AI could have significant implications for jobs, skill demand, and AI adoption in various industries.

By making AI models freely available, more companies will access and experiment with the technology, which may fuel AI-startups and job creation.

As AI becomes more accessible, there should be a greater demand with professionals with open-source AI skills. Additionally, more research-oriented roles may be created in academic organizations, and if its partnerships continue,new specialized roles could emerge in AI integration and optimization. It’s early days and all to play for.

3 high-paying roles to apply for today

• Senior DevSecOps Engineer, SciTec, Boulder ($125,000 - $168,400)
• Sr R&D Engineer, Lowe's, Kirkland ($163,800 - $311,200)
• Senior Software Engineer - BackEnd, Vorto Operations LLC, Denver ($100,000 - $200,000)

Collaboration in open-source

Zuckerberg says he believes AI should and will develop in a similar way to Linux, but currently, most big tech companies are leading the charge with closed models.

But really, the open letter was drafted to announce the latest Llama model, which – surprise, surprise – is open source.

He wrote: “Today we're taking the next steps towards open source AI becoming the industry standard. We're releasing Llama 3.1 405B, the first frontier-level open source AI model, as well as new and improved Llama 3.1 70B and 8B models.

“In addition to having significantly better cost/performance relative to closed models, the fact that the 405B model is open will make it the best choice for fine-tuning and distilling smaller models.”

As well as releasing these models, he announced that Meta is also collaborating with a range of companies to grow the broader ecosystem.

Amazon, Databricks, and Nvidia are all launching comprehensive suites of services to support developers in fine-tuning and distilling their models.

Meanwhile, companies like Scale.AI, Dell, and Deloitte are ready to help enterprises adopt Llama, and train custom models with their own data.

Benefits of open source for developers

Zuckerberg says open-source AI allows developers to take advanced models, continue training them with their own data while ensuring it meets their specific requirements –  all without relying on external parties.

From a cybersecurity perspective, open source means data isn’t shared beyond a company and over cloud APIs, reducing risk and increasing security.

He also gives the hard sell that developers may execute inference on Llama 3.1 405B on their own infra at about half the cost of using closed models like GPT-4o.

Ultimately, a big issue with a closed-model provider of anything is that it can change its model, terms of use, or even cease offering the product completely at any time. While open source means companies aren’t at the mercy of another organization’s business strategy.

Zuckerberg says his personal experiences, such as the constraints faced when building services on Apple's platforms, have reinforced his belief in open ecosystems.

3 great tech jobs hiring now

• Data Center Engineer, Paycom Payroll Llc, Oklahoma City
• Java Developer, The Hollister Group, Boston
• Principal Quality Assurance Architect, Options Clearing Corporation, Chicago

Benefits of open source for Meta

Zuckerberg goes on to list the benefits for Meta, and very much leads with philosophical sentiments.

But as well as reiterating his viewpoint that open ecosystems will build the next generation of computing, he also states that AI models aren’t Meta’s business model so it can afford to go open source.

He also mentions that having ecosystem innovators standardize its open-source tools, like PyTorch and React, has been beneficial for the organization in the past.

Essentially, Meta can benefit from community improvements and innovations without bearing all the development costs or responsibility for potential misuse themselves.

Cynics might also join some dots by considering Meta’s failure to dominate VR with the Metaverse, it is now deciding to hurt its competitors, like Apple and Google, by democratizing AI. Both companies are name-dropped in the letter.

Zuckerberg explicitly mentions his frustration with Apple's constraints on Meta's services, while his mention of Google is during his section Meta’s safety process.

By stating that the information is "already on the internet" and can be "quickly retrieved from Google or other search results," Zuckerberg seems to be implying that if harmful information is accessible through Meta’s AI, it's not uniquely its fault.

Additionally, widespread adoption of its AI models could not only give Meta a foothold in the industry, but could also provide them with valuable insights into how AI is being used across industries. Again, furthering a competitive advantage.

Finally, by highlighting how several closed-source providers lobby government about open source, and by advising “how the US and democratic nations should handle the threat of states with massive resources like China”, the Meta founder furthers his positioning of open-source AI as more transparent.

Zuckerberg says open source gives the US, and its allies, a longer-term advantage by democratizing the technology to “startups, universities, and small businesses”, and recommends that the best strategy is leading companies working closely with the government and its allies. This could be interpreted as a strategy to preempt stricter government regulation of AI.

Ultimately, Zuckerberg’s Llama 3.1 announcement is not just a declaration of Meta’s strategy but a call to action for the entire tech community. His clear bullet-pointed list of ‘Why open source AI is good for developers’ shows that he is keen to get the community on board. But whether Llama becomes the next Linux-like success story remains to be seen…

Ready to find your next role in tech? Whether you want to pivot to AI or put your coding expertise to good use, visit the Hackernoon Job Board today