Elon Musk’s company may also support a Hungarian proptech startup
Elon Musk’s IT company, OpenAI—the founder of Tesla and SpaceX—developed GPT-3, the world’s most advanced artificial intelligence to date, to which access is granted gradually and only to a limited number of developer companies worldwide. Now, Hungary’s Realmonitor Kft. has gained access to this technology. The company helps buyers find the best offers for their chosen properties.
How does GPT-3 help in the real estate market?
Artificial intelligence is becoming embedded in more and more aspects of life and has already become a core element of most services. It powers personalized offerings from tech giants, Google searches, Netflix’s evening movie recommendations, helps cars optimize fuel consumption, and even assists researchers in developing new medicines.
The “sophistication” of AI is measured by the number of parameters used in its training. The more trainable parameters an AI has, the “smarter” it is. With technological progress, this number has grown rapidly—from a few dozen or thousand initially, to 150 million in recent years.
At this point, AI already easily surpasses humans in chess, Go, and poker, performs better in quiz shows, and often diagnoses patients more accurately than doctors. Elon Musk’s company, OpenAI, has now taken another leap by creating GPT-3, the most advanced AI to date, with no less than 175 billion parameters.
This AI can already write articles at human level, conduct natural conversations, translate at a native level, and demonstrate abilities not even known to its creators.
Árminimum: ending the chaos of property listings
Precisely because of its unknown abilities and potential misuse, GPT-3 is only gradually being made available to technology companies. Realmonitor Kft., operator of arminimum.hu, is among the first in Hungary to gain access to this AI, in order to help families find homes more effectively.
With GPT-3, home seekers can view all available listings for a given property at once—even if each ad uses different images, parameters, or significantly different prices for the same property.
How much can AI save in the real estate market?
The significance of this is highlighted by the fact that the “chaos index” of the Hungarian real estate market is 23%—meaning that 23% of online listings actually duplicate the same property, often at a much lower price elsewhere. According to Árminimum’s survey, the average price difference for properties listed multiple times is HUF 4 million when buying, while renters can save an average of HUF 37,000 per month on the same property if they research more thoroughly.
