Lifeboat Foundation

Scientists discovered a large repository of brain genes in a sea sponge. Why would an ancient, porous blob of cells contain neural genes?

Benchmarks are a key driver of progress in AI. But they also have many shortcomings. The new GPT-Fathom benchmark suite aims to reduce some of these pitfalls.

Benchmarks allow AI developers to measure the performance of their models on a variety of tasks. In the case of language models, for example, answering knowledge questions or solving logic tasks. Depending on its performance, the model receives a score that can then be compared with the results of other models.

These benchmarking results form the basis for further research decisions and, ultimately, investments. They also provide information about the strengths and weaknesses of individual methods.

Chipotle announces that robots will soon be assembling bowls and salads.

Robots are great specialists, but poor generalists. Typically, you have to train a model for each task, robot, and environment. Changing a single variable often requires starting from scratch. But what if we could combine the knowledge across robotics and create a way to train a general-purpose robot?

Today, we are launching a new set of resources for general-purpose robotics learning across different robot types, or embodiments. Together with partners from 33 academic labs we have pooled data from 22 different robot types to create the Open X-Embodiment dataset. We also release RT-1-X, a robotics transformer (RT) model derived from RT-1 and trained on our dataset, that shows skills transfer across many robot embodiments.

In this work, we show training a single model on data from multiple embodiments leads to significantly better performance across many robots than those trained on data from individual embodiments. We tested our RT-1-X model in five different research labs, demonstrating 50% success rate improvement on average across five different commonly used robots compared to methods developed independently and specifically for each robot. We also showed that training our visual language action model, RT-2, on data from multiple embodiments tripled its performance on real-world robotic skills.

A battery built for satellites brings grid-scale storage down to Earth.

In what may be one of Earth’s craziest forms of mimicry, researchers have discovered a new species of rove beetle that grows a termite puppet on its back to fool real termites into feeding it. The replica is so precise, it even mirrors the termites’ distinct body segments and has three pairs of pseudoappendages that resemble antennae and legs.

Rove beetles (family Staphylinidae) are already infamous in the animal kingdom as masters of disguise. Some, for example, have evolved to look like army ants, allowing the beetles to march alongside them and feed on their eggs and young.

The new beetle species (Austrospirachtha carrijoi)—found beneath the soil in Australia’s Northern Territory—emulates a termite by enlarging its abdomen, a phenomenon known as physogastry. Evolution has reshaped this body part into a highly realistic replica of a termite (as seen above), head and all, which rides on top of the rest of the beetle’s body. The beetle’s real, much smaller head peeks out from beneath its termite disguise, the authors report this month in the journal.

I WOULD SUSPECT IF WE CAN GET OFF THIS PLANET AND FIND SOMEWHERE MORE HABITABLE, THAT WOULD BE MORE PREFERABLE.

The world’s land masses are, a new super-charged climate model suggests, going to form into one giant supercontinent — and if humans manage to survive the shift, we will become like the inhabitants of Arrakis, the desert planet at the heart of the “Dune” series.

A new study led by researchers at the University of Bristol and published in the journal Nature Geosciences predicts that over the next 250 million years, the continents will shift to form what they’re calling “Pangea Ultima,” an uber-hot supercontinent that will be inhospitable to most mammals due to the conditions that made it.

Continue reading “Scientists Say Earth Will Become a Barren Wasteland” »

This is a risky bet, given the limitations of the technology. Tech companies have not solved some of the persistent problems with AI language models, such as their propensity to make things up or “hallucinate.” But what concerns me the most is that they are a security and privacy disaster, as I wrote earlier this year. Tech companies are putting this deeply flawed tech in the hands of millions of people and allowing AI models access to sensitive information such as their emails, calendars, and private messages. In doing so, they are making us all vulnerable to scams, phishing, and hacks on a massive scale.

I’ve covered the significant security problems with AI language models before. Now that AI assistants have access to personal information and can simultaneously browse the web, they are particularly prone to a type of attack called indirect prompt injection. It’s ridiculously easy to execute, and there is no known fix.

In an indirect prompt injection attack, a third party “alters a website by adding hidden text that is meant to change the AI’s behavior,” as I wrote in April. “Attackers could use social media or email to direct users to websites with these secret prompts. Once that happens, the AI system could be manipulated to let the attacker try to extract people’s credit card information, for example.” With this new generation of AI models plugged into social media and emails, the opportunities for hackers are endless.

Sure, you could just stick a ChatGPT sidebar in your browser. But what do we really want AI to do for us as we use the web? That’s the much harder question.

At some point, if you’re a company doing pretty much anything in the year 2023, you have to have an AI strategy. It’s just business. You can make a ChatGPT plug-in. You can do a sidebar. You can bet your entire trillion-dollar company on AI being the future of how everyone does everything. But you have to do something.

The last one of these was crypto and the blockchain a couple of years ago, and Josh Miller, the CEO of The Browser Company, which makes the popular new Arc browser, says he’s… More.

Continue reading “AI is coming to the Arc browser — but probably not like you think” »