Mass produced humanoid robots are coming, the U.S. government is warning NVIDIA not to ship AI chips to China. HP’s printer lock-in strategy is unpopular but profitable and is it time to rethink the Turing test?
These and more top tech stories on Hashtag Trending
I’m your host Jim Love, CIO of IT World Canada and Tech News Day in the US.
Agility Robotics is set to open a factory in Salem, Oregon, with the ambitious goal of mass-producing 10,000 bipedal robots annually.
This facility, named RoboFab, will be the first to mass-produce humanoid robots, potentially revolutionizing industries like Amazon that require heavy lifting and moving tasks. The robots, named Digit, are designed to be nimbler and more versatile than traditional industrial robots.
Digit, a human-shaped robot optimized for warehouse work, has already garnered a growing backlog of orders. Agility Robotics plans to initially produce hundreds of these robots, eventually ramping up to their full capacity. Companies participating in Agility’s Partner Program will receive their robots in 2024, with broader delivery expected in 2025.
Building bipedal robots that can walk without falling and work safely alongside humans is a significant engineering challenge. Competitors in this space include Tesla with Optimus, Boston Dynamics with Atlas, and several others.
Amazon has shown particular interest in Digit, testing the robot at a lab near Seattle for tasks like tote recycling. Amazon’s chief roboticist, Tye Brady, highlights the robot’s potential for navigating uneven surfaces and tight corners, which could be advantageous in human-designed spaces.
This development predicts a growing demand for humanoid robots. We have one question. Will the workers in these factories be robots?
Sources include: Axios
NVIDIA has received a warning from U.S. Commerce Secretary Gina Raimondo regarding its efforts to redesign chips for the Chinese market. This warning comes in response to NVIDIA’s attempts to reportedly circumvent the recent U.S. export restrictions on advanced chips.
NVIDIA is reportedly preparing a new series of HPC (High-Performance Computing) and gaming GPUs to overcome regulatory challenges.
However, Raimondo cautioned that any new product representing a redesign of a previously prohibited processor would immediately become subject to additional restrictions. She emphasized that if a chip is redesigned to enable AI capabilities, it will be controlled by the U.S. government.
This situation highlights the ongoing tension between U.S. export controls and technology companies’ efforts to adapt to a rapidly changing global market.
Sources include: VideoCardz
And here’s another story of struggles with foreign standards. You remember that Apple has been forced to go to USB-C for its new models in Europe as the government has demanded a “single charger rule” for phones.
India has gone one step further and demands that Apple update prior models as well. This is significant because a great deal of Apple’s sales are from older models of the iPhone.
India’s rules affect all electronic devices sold in the country, including those not originally designed with USB-C.
During a closed-door meeting with India’s IT ministry on November 28, Apple requested exemptions for some older iPhone models. The company expressed concerns that incorporating USB-C into older models would hinder its ability to meet the Production-Linked Incentive (PLI) targets set by Prime Minister Narendra Modi’s major project, which offers financial incentives for investments and incremental annual phone sales.
Apple proposed a compliance timeline of June 2025 if existing models are exempted, but it would need an additional 18 months beyond 2024 without an exemption.
Sources include: AppleInsider
Speaking at the UBS Global Technology conference, HP’s chief financial officer, Marie Myers highlighted that moving customers from a transactional model to a subscription model like Instant Ink, which includes paper, increases the value of a customer by about 20 per cent. This model essentially “locks in” customers into a longer-term relationship.
Instant Ink operates by sending ink or toner cartridges to customers as needed, with plans ranging from 99 cents to $25.99 per month. Despite its high-end pricing, the service had over 11 million subscribers as of May last year. HP has faced criticism for its approach to printer hardware and ink cartridges. In 2019, the company increased the price of printer hardware and shifted focus to Smart Tank and Neverstop printers, pre-loaded with an estimated two years’ worth of ink or toner.
HP has been criticized for blocking third-party ink cartridges in its printers, citing its dynamic security policy introduced in 2016. This policy aims to protect HP’s intellectual property and ensure customer experience quality by only allowing the use of cartridges with new or reused HP chips or electronic circuitry.
Despite the backlash, these controversial policies have proven financially beneficial for HP, with its printing division margin increasing from 14.8 per cent in fiscal 2020 to 18.9 per cent in fiscal 2023.
Sources include: TechSpot
In 2023, the most popular topic on Wikipedia’s English-language site was the article describing ChatGPT, the AI system developed by OpenAI.
Wikimedia Foundation, annually releases a list of the most-viewed articles on the platform. The “ChatGPT” article received the highest number of page views, totaling 49.5 million. Throughout the first half of the year, the article consistently attracted attention, with daily page views ranging between 100,000 and 400,000.
The top 10 list of most-viewed articles on Wikipedia reflects diverse interests, including four articles on cricket and two on Bollywood movies, highlighting the significant English-speaking population in India. For the first time, cricket made the list, with 38 million views, a 304 per cent increase from the previous year. The Indian Premier League also featured prominently with 32 million views.
Taylor Swift emerged as the most popular celebrity on Wikipedia in 2023, with her page receiving about 19.5 million views. Other notable figures with high page views included the late Matthew Perry, Lisa Marie Presley, and Elon Musk.
Sources include: Axios
In a recent Turing test study titled “Does GPT-4 Pass the Turing Test?”, researchers from UC San Diego explored the ability of various AI models, including OpenAI’s GPT-4, GPT-3.5, and the 1960s program ELIZA, to convince participants they were human.
Surprisingly, the study found that human participants correctly identified other humans as human in only 63 per cent of the interactions. Even more intriguing, ELIZA outperformed GPT-3.5 in this test.
The Turing test, conceived by Alan Turing in 1950, is a benchmark for determining a machine’s ability to imitate human conversation. In this study, human judges interacted with either another human or a chatbot without knowing which was which. The success rate of the chatbot in convincing the judge of its humanity determines if it has passed the test.
The study involved 652 participants who completed 1,810 sessions. ELIZA achieved a 27 per cent success rate, while GPT-3.5 scored 14 per cent.
GPT-4 achieved 41 per cent, second only to actual humans.
GPT-3.5’s poor performance is partly attributed to its conditioning by OpenAI not to present itself as human.
The study’s findings raise questions about the effectiveness of the Turing test as a measure of AI model performance and the strategies used by humans to detect AI.
Sources include: Ars Technica
And that’s what’s trending today.
Hashtag Trending goes to air 5 days a week with a special weekend interview show we call “the Weekend Edition.”
You can get us anywhere you get audio podcasts and there is a copy of the show notes at itworldcanada.com/podcasts
I’m your host Jim Love. Have a wild and wonderful Wednesday.
>>> Read full article>>>
Copyright for syndicated content belongs to the linked Source : ITBusiness.ca – https://www.itbusiness.ca/news/hashtag-trending-dec-6-mass-produced-humanoid-robots-coming-soon-u-s-government-warns-nvidia-not-to-ship-ai-chips-to-china-time-to-rethink-the-turing-test/126756
