Open Source: Why Corporate Powerhouses Can’t Control Chatbots
Here’s the secret, buried in technobabble within this report written by an engineer at Google: the monopolists have already lost! You don’t have to understand all the jargon to get the point: Today’s chat universe is already out from under the grip of corporate control. And it isn’t going back into the box. The report compares Google with Open AI (a major for-profit competitor and the creator of ChatGPT) and with new open-source chat and AI (artificial intelligence) tools. The report concludes that the open source tools are faster, more nimble, and much faster to deploy and train, despite far less access to resources than projects at companies like Google and Microsoft:
Open-source models are faster, more customizable, more private, and pound-for-pound more capable. They are doing things with $100 and 13B params [13 billion parameters] that we struggle with at $10M and 540B. And they are doing so in weeks, not months.
This has enormous ramifications in every aspect of society. Private profit will not be a driver toward centralization of control and limiting who can play. And that will drive enormous innovation, in ways that I at least can’t yet imagine, let alone describe. Some of it will be monetizable, just as Red Hat monetized an open source operating system (Linux) and Google has monetized its no-charge search engine. But almost none of it will be controllable.
If you’re an investor, know that you can’t buy your way into control–but you can invest in promising developments that will leapfrog to higher good on the basis of freebie systems. And knowing that going in, you’re much less likely to get burned. You won’t, for instance, make the enormous mistake that Rupert Murdoch’s News Corp did when it bought MySpace (at the time, the dominant proprietary social network), spending $580 million on something that became relatively worthless as soon as Facebook began to hit its stride; News Corp sold it just six years later, to Justin Timberlake, for a mere $35 million. That was a loss of more than half a billion dollars.
Let’s look at a few examples of how public intellectual property has impacted us in the past:
- Jonas Salk refused to monetize his polio vaccine and made it widely available. 3000 children died of polio in 1952, and 60,000 were infected (about a third of whom became paralyzed) Polio cases dropped from 21,000 paralytic cases in 1952 to just 61 in 1965. The IPV vaccine was introduced in 1955 and the oral vaccine in 1961.
- Tim Berners-Lee invented the World Wide Web and put it in the public domain. Had he tried to keep it private, most of the open information society and global ecommerce we enjoy today probably would not exist–and especially not the millions of websites built on open source platforms such as WordPress. WordPress powers 43 percent of all websites, and 65 percent of those using any sort of content management system. Shopify, with, um, six-and-a-half percent, is in second place, with the next three all getting less than 3% market share apiece. Even though he gave the Web away, Berners-Lee’s net worth of $50 million is plenty to live on. Anyone out there still paying $20 per month plus usage fees to access the Internet over Compuserve?
- TED Talks were once available only to attenders at the organization’s pricy conferences. But then they decided to open up the content, online, at no charge. This was so successful that TED Talks (and later, TEDx Talks all around the world) became integrated into the wider culture, democratizing the spread of world-changing ideas.
Where might it take us next? I’m not a futurist and I won’t get super-specific, but I do see a few general trends likely to arise from this:
- Just as TV and Internet have converged, so will AI and 3D printing: a combination that will revolutionize tangible goods with innovations in design, manufacturing, and localization
- AI could lead to the next miniaturization revolution, making it easier to do things like holographically project keyboards and maybe even monitors that make it possible to use watch-sized computers and phones a lot more easily
- With the right protocols in place, AI could do a lot of basic research and interpretation of the data o that students could concentrate on learning concepts and following them down unexpected paths to create new concepts–OR, without those protocols, it could be used to reduce learning to something meaningless
- In a perfect world–and we can help it become more perfect–AI can help us understand and solve our most pressing problems. It could turn us from linear to circular resource use, where every output becomes not waste, but something a different process could use. It could redistribute food, housing, and shelter (and other resources) so everyone has enough and no one has 5 million times too much, using only natural materials, generating zero waste, using zero net energy, and creating zero pollution.
Regardless of how it turns out, expect any sector you’re in–for example, agriculture, manufacturing, distribution, transportation, services, tech, creative arts, nonprofit, academia, government, military, or whatever–expect your world to turn upside down.