Category: Uncategorized

VIGILANCE REQUIRED. “There are fewer obvious constraints on the harm a system with poorly specified goals might do. In particular, an autonomous system that learns about human goals, but is not correctly designed to align its own goals to its best model of human goals, could cause catastrophic harm in the absence of adequate checks.” https://lnkd.in/dWsky-W View in LinkedIn

THE RISK. Systems that end up with poor models of their environment can do significant harm. View in LinkedIn

AMPLIFIED CONSEQUENCES. “As perception, inference, and planning algorithms improve, AI systems will be trusted with increasingly complex and long-term decision-making. Small errors will then have larger consequences.” https://lnkd.in/dWsky-W View in LinkedIn

PREDICTIONS from the Machine Intelligence Research Institute (MIRI) follow. ( Stuart Russell, Advisor). https://lnkd.in/dWsky-W View in LinkedIn

MICROCHIPS. "It’s a messy soup of electrons, and the supposedly binary values of one and zero (or on and off) are not the only option in town. (Because of the vagaries of voltage, a logic gate, Smart explains, can effectively be stuck on “half”.)" https://lnkd.in/dJ7JeAs View in LinkedIn

Guess View in LinkedIn

DEMOCRATIC SCIENCE. “While artificial intelligence (AI) sounds esoteric to most, their goal is really quite simple: to democratize AI technology, a complex world previously only accessible to data scientists (a few thousand worldwide). Now, any of the world's 18 million developers can incorporate AI technology into their apps and programs, allowing them to build smarter and faster.” https://lnkd.in/dqkvFim View in LinkedIn

ENCRYPTION. “Google Brain’s neural-net AI dreams up its own encryption strategy; neural networks can learn to protect the confidentiality of their data from other neural networks; the day may one day come when two computers working together can self-generate a custom encryption scheme more robust than anything humans can imagine, then discard within minutes and create a new one as a form of information security.“ https://lnkd.in/dvWV-3G View in LinkedIn

BLACK BOX. “In the past, Google relied mostly on algorithms that followed a strict set of rules set by humans. The concern—as described by some former Google employees—was that it was more difficult to understand why neural nets behaved the way it did, and more difficult to tweak their behavior. The truth is that even the experts don't completely understand how neural nets work.” https://lnkd.in/d-4Nkau View in LinkedIn