Archive 26.06.2024

Researchers have created a new, cost-effective training technique for artificial intelligence systems that aims to make them less socially biased.

Computers have come so far in terms of their power and potential, rivaling and even eclipsing human brains in their ability to store and crunch data, make predictions and communicate. But there is one domain where human brains continue to dominate: energy efficiency.

Researchers have found a way to bind engineered skin tissue to the complex forms of humanoid robots. This brings with it potential benefits to robotic platforms such as increased mobility, self-healing abilities, embedded sensing capabilities and an increasingly lifelike appearance. Taking inspiration from human skin ligaments, the team included special perforations in a robot face, which helped a layer of skin take hold.

The way consumers find information through search engines is changing again — this time, with the introduction of Generative AI into search results. Google and other search engines use Generative AI to give individuals simplified search results. These AI-powered results […]

The post Adapting to AI in Search: How Local Businesses Can Prepare for Customer Behavior Changes appeared first on TechSpective.

Researchers have found a way to bind engineered skin tissue to the complex forms of humanoid robots. This brings with it potential benefits to robotic platforms such as increased mobility, self-healing abilities, embedded sensing capabilities and an increasingly lifelike appearance.

A force sensor measures mechanical forces, like a secret handshake between machines and the real world. These ingenious devices convert mechanical force into electrical signals, enabling us to measure, analyze, and interact with the physical world around us.

As sub-symbolic AI, like deep learning, continues to advance, its limitations in safety and reliability are becoming more apparent. Verification and stability are crucial in safety-critical domains such as humanoid robotics, which is rapidly evolving into a versatile tool for various applications. However, proving the correctness of AI-based self-learning algorithms is challenging due to their uncertain inferences and opaque decision-making processes.

Research on energy efficiency enhancement of a tortoise-inspired legged robot, by a research team led by Dongwon Yun, professor at the Department of Robotics and Mechatronics Engineering at DGIST, has been featured on the cover of Advanced Intelligent Systems.

One of the troubling uses of a new technology like AI is the feeding frenzy of trying to apply it to every problem without any significant effort to inform the order in which AI is implemented by company need. This […]

The post IBM watsonx Assistant for Z: Applying AI Where the Need Is Greatest appeared first on TechSpective.

The artificial intelligence (AI) boom has picked up pace over the past two months, with enterprise companies around the world scrambling to integrate the latest breakthroughs. What once seemed like a far-off vision is now an imminent reality – AI […]

The post Enterprise AI Adoption Explodes After the Major Breakthroughs of 2023 appeared first on TechSpective.

Meet CARMEN, short for Cognitively Assistive Robot for Motivation and Neurorehabilitation -- a small, tabletop robot designed to help people with mild cognitive impairment (MCI) learn skills to improve memory, attention, and executive functioning at home.

Meet CARMEN, short for Cognitively Assistive Robot for Motivation and Neurorehabilitation—a small, tabletop robot designed to help people with mild cognitive impairment (MCI) learn skills to improve memory, attention, and executive functioning at home.

Motion control optimization is about more than ensuring a robot can move. It must also move correctly; meeting the ideal speed, power, direction, length and timing. Assurance in each of these categories is necessary because these are what separate robots from human employees

AI Reduces 60+ Copywriting Team to One Editor

In yet another bone-chilling example of how AI is hollowing-out copywriting teams, this BBC report details how AI turned a 60+ copywriting team into a one-man operation.

First introduced by the publisher in 2023, AI slowly began to usurp more and more jobs until by 2024, everyone on the team was vaporized save for one, lone editor.

Observes the last of the team, who chooses to remain anonymous: “All of a sudden, I was just doing everyone’s job.

“Mostly, it was just about cleaning things up and making the writing sound less awkward, cutting out weirdly formal or over-enthusiastic language.

“It was more editing than I had to do with human writers, but it was always the exact same kinds of edits. The real problem was it was just so repetitive and boring. It started to feel like I was the robot.”

That account is a long way from current-day AI evangelism, which insists AI is little more than a warm-and-fuzzy friend who will always help you — and never hurt.

For editors and writers who are not tasked with unearthing fresh news data in their jobs, the message is clear: Increasingly, staying alive in copyediting has become a fight to be ‘the last one standing.’

In other news and analysis on AI writing:

*In-Depth Guide: New AI Writer Challenger: Close Enough to Make ChatGPT Yawn: Reviewer Jayric Maning finds that while that Llama3 AI chatbot is no slouch, it still comes in behind market leader ChatGPT.

Observes Maning: “I would have to say that GPT-4 is the better LLM (AI engine). GPT-4 excels in multimodality, with advanced capabilities in handling text, image, and audio inputs, while Llama 3’s similar features are still in development.

“GPT-4 also offers a much larger context length and better performance and is widely accessible through popular tools and services, making it more user-friendly.”

*ChatGPT’s New Smarts: Maybe Humans Should Stick to Finger-Painting?: While knowledge of ChatGPT’s intellectual prowess is widespread, new metrics indicate the AI writer/tool is smarter than many realize.

Specifically, ChatGPT’s performance on numerous standardized tests — compiled by former researcher OpenAI Leopold Aschenbrenner — leaves most humans hopelessly behind.

Here’s how ChatGPT ranks against humans on some of the most common — and challenging — high school and college exams, according to Aschenbrenner:

~Uniform Bar Exam: Top 10%

~LSAT (Law School Admission Test): Top 12%

~SAT (Scholastic Aptitude Test): Top 3%

~GRE (Graduate Record Examination) (Verbal): Top 1%

~GRE (Graduate Record Examination) (Quantitative): Top 20%

~US Biology Olympiad (high school): Top 1%

~AP Chemistry(high school): Top 20%

~AP Macroeconomics (high school): Top 8%

~AP Calculus BC (high school): Top 49%

One caveat: ChatGPT took the above tests in 2023, before it was upgraded to ChatGPT-4o.

So chances are, ChatGPT-4o would score even higher on some or all of these tests above.

*Anthropic’s New Claude 3.5: An Ego-Check for ChatGPT?: In a win for consumers, Anthropic has released an upgrade to its ChatGPT competitor that it claims outperforms ChatGPT.

While the jury is still out, Anthropic claims Claude 3.5:

~Writes in a more natural tone

~Is better at nuance and humor

~Processes complicated prompts more easily

Either way, writers win: Brilliant AI programmers at various AI writing firms remain hell-bent on outdoing each other.

And consequently, the auto-writing tools just keep getting more sophisticated.

*Brand Writing, Steadfastly Consistent: Acrolinx — an editing and writing tool designed to ensure all users write in the same brand voice — has gotten an AI upgrade.

The changes enable users to instantly auto-generate ‘brand-standard’ writing at a pre-set quality level.

It also double-checks any writing auto-generated to ensure it steers clear of plagiarism.

*New WordPress Plug-In: Automated Writing, with a Side of SEO: WordPress users now have another AI writer optimized for scoring high in search engine returns: ‘SEO Basics — AI Writer.’

In addition, the tool also enables the creation of automated posts that can be released on a scheduled basis.

Plus, it can transcribe YouTube videos to written transcriptions and includes social media filters that enable ‘media-rich content that aligns with current trends and topics.’

*Snapshot: Some Top AI Paraphrasing Tools: Business Research Insights has released its list of top AI paraphrasers. Here’s their selection, along with links to pricing pages:

~CoderDuck

~SEO Wagon

~Spin Rewriter

~Quillbot

~Prepost SEO

One caveat: Virtually all AI writers — including ChatGPT, Google’s Gemini and Anthropic’s Claude — are capable of paraphrasing.

AI tools specializing in paraphrasing are pitched as ideally designed to make AI paraphrasing easier and more powerful.

*ChatGPT: Picking Your Favorite Flavor: Writer Rachel Davies offers an excellent rundown in this piece of the latest versions of ChatGPT and their costs.

The quick takeaway: Besides ChatGPT’s limited free version, ChatGPT Plus is available for $20/month, ChatGPT Team runs $30/month and ChatGPT Enterprise costs $60/month.

*Bad Dog: Adobe Sued by U.S. Over Tough-to-Unsubscribe Tricks: Consumers fooled by tricky, bait-and-switch, ‘easy unsubscribe’ offers will most likely cheer this day in court against Adobe.

The beef: The U.S. alleges that Adobe — provider of a number of AI writing/imaging services — deliberately misled customers about how to cancel its subscriptions.

Observes writer David McCabe: “Adobe took steps to lock consumers into yearly subscriptions billed in monthly increments, the lawsuit argued.

“The overall price of the plan was often displayed in bold when customers signed up.

“But a reference to Adobe’s cancellation fee was displayed in lighter italic text, the government said.

“Consumers had to click a separate link to see details of the early cancellation fee, which cost half of any remaining payments and applied if a customer canceled in the first year, the government said.”

*AI Big Picture: Sad Masquerade: CNN Exposes AI Posing as Human Reporters: Writer Hadas Gold takes an in-depth look in this piece at a growing problem with AI writing: Publishers who mislead readers that AI-generated articles are being written by human beings.

The company in the crosshairs in Gold’s piece: Hoodline.

Observes Peter Adams, a senior vice president of the News Literacy Project: The way the site uses and discloses AI purposely tricks readers by “mimicking” the look and feel of a “standards-based local news organization with real journalists.”

Share a Link:  Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.

–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.

The post The Myth of the ‘Cheery, AI Collaborator’ appeared first on Robot Writers AI.

An elephant uses its trunk for eating, drinking water, communicating, exploring the environment, social behavior, and making and using tools. The trunk, which contains six muscle groups, is not only very strong—it can uproot a tree—but can be used with great precision. Elephants use a number of techniques to grasp objects, including suction, pinching with the two "fingers" at the tip of the trunk and wrapping the trunk around the object.