AGI starts here

INVEST NOW… (coming soon)

Rooted on our discoveries, we are adding new layers of cognitive elements (named neural abstractions), their internal relationships and mathematics; so machines can comprehend mass, time and space (and a few more things); resulting in a single-instant learning, eco-friendly and fully reasoning intelligence (an AGI, not an AI) that doesn’t suffer from hallucinations, nor requires the massive data, repetitive training, nor enormous power of AI. Instead, with an AGI, a regular laptop can learn, understand, reason and think like we do.

In software, we are building powerful expert minds (AGI agents) that can be installed in any computer, and continue learning the specific needs of every user, to solve the problems and deliver results that AI cannot.

AI, like ChatGPT, needs trillions of sentences (aka training) to memorize and repeat the sequence of words. So, if the machine doesn’t have the next word in record, it guesses, likely getting it wrong (aka hallucinations). However, an AGI like Abstra, doesn’t memorize/repeat words. Instead, like you and me, it comprehends what words say. This means that with a single sentence and in fractions of a second, the machine can reason, act, answer and respond. Indeed, intelligence could be measured by the lowest number of iterations (the more repetitions, the less intelligent one is)

In a hybrid scenario, AI/LLMs can use our AGI, Abstra, to keep their answers grounded in reality to avoid hallucinations and unnecessary words.

In version #1 (comprehension of mass), we authored 125 US patents. In our latest version #2 (comprehension of mass, time, space & others), our new intellectual property is close to x10 in size.