Scale is one of the biggest challenges facing the AI and Robotic “Revolution,” how will you inform all the infrastructure needed for all these autonomous tools? One answer you’ll hear a lot more often: crowds. While OpenAI and other ultra-intelligent Firsts continue to build multi-billion dollar data centers, a more automated approach is starting to catch on.
The Register recently highlighted one company that is taking the “Swerm” approach to AI, and it’s apparently producing some impressive results. The start-up company FortyTwo has published benchmarks that show how to use small AI models on your computers that have been able to deal with consulting models, seeking better results than OPPT-5 of Claude Opus 4.1, and R1 of ReveSeed’s Claude Opus 4.1, and R1 of ReveSeek’s.
The idea of the company, which seems to have exploded results, is very simple: large aa models can get stuck in showing loops when given complex tasks. Small models, meanwhile, are not only the limit that reflects the mortality but also provide many answers that can be calculated to get a better answer. That, in theory, is done by distributing computing operations across multiple devices rather than in a single, centralized data center – not that cryptocurrencies plow operations across multiple devices. Accordingly, this company offers a crypto-based reward
The same is developing in the robotic realm, with a backstreet magazine recently highlighting research showing how robots work with a form of interactive communication. The researchers suggested that such a model could be used in settings such as fire monitoring, where drones are monitored for potential fires. If one DROWE were to fail, the others would come down to pick up the slack. Another example is floating researchers are growing small robots that can’t interact with each other to do things like deliver drugs inside the human body or perform surgeries. The latter apparently wasn’t theoretical: The researchers used tiny, chain-push robots as a unit to clear blockages in artificial blood vessels.
The basic idea behind the research is that these machines can be very simple but perform complex tasks when working together. The researchers completed a test in which a group of robots that were able to perform three skills – move forward, follow sound, and listen to neighbors – were able to connect obstacles as best they could. Apparently, the old idiom, “two heads are better than one,” applies to AI and robots, too.
