Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 1,800

Tesla Autopilot tries valiantly before asking for help in Vietnam’s crowded streets

In the future, Tesla’s Autopilot and Full Self-Driving suite are expected to handle challenging circumstances on the road with ease. These involve inner-city driving, which includes factors like pedestrians walking about, motorcyclists driving around cars, and other potential edge cases. When Autopilot is able to handle these cases confidently, the company could roll out ambitious projects such as Elon Musk’s Robotaxi Network.

Tesla’s FSD Beta, at least based on videos of the system in action, seems to be designed for maximum safety. Members of the first batch of testers for the FSD Beta have shared clips of the advanced driver-assist system handling even challenging inner-city streets in places such as San Francisco with caution. But even these difficult roads pale in comparison to the traffic situation in other parts of the world.

In Southeast Asian countries such as Vietnam, for example, traffic tends to be very challenging, to the point where even experienced human drivers could experience anxiety when navigating through inner-city roads. The same is true for other countries like India or the Philippines, where road rules are loosely followed. In places such as these, Autopilot still has some ways to go, as seen in a recently shared video from a Tesla Model X owner.

BrainGate: First Human Use of High-Bandwidth Wireless Brain-Computer Interface

Summary: The BrainGate brain-machine interface is able to transmit signals from a single neuron resolution with full broadband fidelity without physically tethering the user to a decoding system.

Source: Brown University.

Brain-computer interfaces (BCIs) are an emerging assistive technology, enabling people with paralysis to type on computer screens or manipulate robotic prostheses just by thinking about moving their own bodies. For years, investigational BCIs used in clinical trials have required cables to connect the sensing array in the brain to computers that decode the signals and use them to drive external devices.

TSMC to Spend $100 Billion Over Three Years to Grow Capacity

In 2020, TSMC spent a record $18 billion on building new factories for their chips. TSMC just announced they are spending $100 billion on new factories over the next 3 years. This will radically change the chip landscape. Many other companies, including Samsung and Intel, are upping their spending as well.

Of course, at some point there will be a chip glut again but this greatly increased chip capacity will change the world that we live in. It will also make AGI (Artificial General Intelligence) that much closer to reality… (All this money gives companies an incentive to spend R&D on smaller transistors, etc.)

Businesses Take a Hurry-Up-and-Wait Approach to AI

But whenever companies experiment with a new technology that has the potential to transform entire business models, like electricity, it can take decades before changes yield real-world results, Mr. Brynjolfsson said, speaking on Wednesday at The Wall Street Journal Pro AI Executive Forum. The Digital Economy Lab is part of Stanford University’s Institute for Human-Centered AI.

Companies leading the charge in adopting AI tools and platforms are taking time to target spending in the right digital capabilities and talent, he said.

“We’re having a few superstars doing really well,” Mr. Brynjolfsson said. “But the whole reason it takes so long in the first place is that it’s not easy.” He expects to see a “productivity J-curve” as companies figure out how best to deploy AI in their daily operations.

Ambi Robotics, formerly Ambidextrous, raises $6.1M for picking systems

Ambi Robotics has two flagship products. AmbiSort is a robotic putwall that sorts boxes, polybags, and envelopes from bulk input flow (chutes, totes, and bins) into destination containers (mail sacks, totes). Ambi Robotics claims the system works “over 50% faster than manual labor.” AmbiKit is a robotic system that builds unique kits from any item set. The company said it can be used with subscription boxes, medical kits, gift sets and sample sets for a variety of industries, including cosmetics, food and beverage, consumer goods, medical devices, aerospace and automotive.

The company’s robots are modular, but they do use suction-based gripping. Here’s how AmbiSort works. A depth-sensing camera first looks into a bin of items and analyzes the objects. After determining how to best grasp the item, the robot picks up the item with its suction gripper, holds it up to a barcode scanner, then places the item into a bin. The system then alerts a human operator when a bin is full and ready to be packed.

Machine Learning Faces a Reckoning in Health Research

Univ. of Toronto Researcher: “I did not realize quite how bad [the lack of reproducibility and poor quality in research papers] was.”

Many areas of science have been facing a reproducibility crisis over the past two years, and machine learning and AI are no exception. That has been highlighted by recent efforts to identify papers with results that are reproducible and those that are not.

Two new analyses put the spotlight on machine learning in health research, where lack of reproducibility and poor quality is especially alarming. “If a doctor is using machine learning or an artificial intelligence tool to aid in patient care, and that tool does not perform up to the standards reported during the research process, then that could risk harm to the patient, and it could generally lower the quality of care,” says Marzyeh Ghassemi of the University of Toronto.

In a paper describing her team’s analysis of 511 other papers, Ghassemi’s team reported that machine learning papers in healthcare were reproducible far less often than in other machine learning subfields. The group’s findings were published this week in the journal Science Translational Medicine. And in a systematic review published in Nature Machine Intelligence, 85 percent of studies using machine learning to detect COVID-19 in chest scans failed a reproducibility and quality check, and none of the models was near ready for use in clinics, the authors say.