This year\u2019s CES show attracted 15% more visitors than last year, with AI taking centre\u00a0stage.<\/p>\n
The event \u2013 which took place at the Las Vegas Convention Center and covered innovations ranging\u00a0from consumer TVs and audio products, through wearables and XR, to industrial automation, enterprise robotics and advancements in automotive \u2013\u00a0welcomed 4,300 exhibitors and 135,000 delegates.<\/p>\n
But the total was still down on pre-pandemic levels, according to Futuresource\u2019s show report, with about 170,000 attending in 2020 and a record 184,729 in 2017.<\/p>\n
AI, meanwhile, is moving beyond \u2018tech demos\u2019 and showcases, and is being positioned for commercial applications and\u00a0real-world products and services.<\/p>\n
However, the Futuresource report notes that it is important to realise that the term AI is being used quite broadly \u2013 it does not necessarily mean intelligence that is integrated within products themselves.<\/p>\n
Generally, AI\u00a0is tasked with analysing\u00a0large data sets and finding patterns that would be hidden or difficult to derive using statistical methods alone. The outcomes are often used to tailor functions on consumer products, allowing them to perform better than they\u00a0would have without AI involvement.<\/p>\n
This is very different\u00a0from machine learning and training of AI, which today mostly resides\u00a0on servers in the cloud \u2013\u00a0although it is steadily migrating into devices. Most companies have now realised\u00a0that the way forward employs the vertical integration of AI.<\/p>\n
Harnessing ChatGPT\u00a0<\/strong> Businesses are training GPT models on specialist data sets that then become expert\u00a0in specific applications \u2013 such\u00a0as\u00a0healthcare or automotive scenarios \u2013\u00a0rather than generic tasks. This is especially important in applications where precision is vital and where AI \u2018hallucinations\u2019 \u2013 where the results look plausible but are incorrect \u2013 cannot be afforded.<\/p>\n At CES, the products on show included\u00a0bathroom mirrors that detect your mood and pillows that claim to reduce snoring, as well as\u00a0cars that integrate ChatGPT and robotic companion devices.<\/p>\n US start-up Rabbit collaborated with tech company Teenage Engineering to launch the Rabbit R1 at CES. The pocket device is designed to navigate various online services to complete tasks, operating\u00a0as\u00a0a virtual assistant but with a more human interface.<\/p>\n Dutch start-up Focus showcased\u00a0a model that aims to predict the next major technology advancements by using AI to dissect and interpret global patent filings.<\/p>\n This \u201cpredictive AI analysis platform\u201d uses AI to forecast not only upcoming trends but also their probability of success, using measures such as the number of patent submissions and the frequency of patent revisions. It also identifies which companies are most active in each technology.<\/p>\n CES also saw the demonstration of silicon chips with dedicated neural processing engines alongside the fast memory interfaces necessary to process vast amounts of data. This has been available on smartphones for several years, but the technology is now targeting smaller microcontrollers for smart cameras and sensors, as well as\u00a0audio and voice processing.<\/p>\n The report notes that marketing and technology experts were excited about how AI will help enable spatial computing. Although the metaverse was practically nowhere to be seen, there were plenty of virtual reality, augmented reality and mixed reality headsets on the show floor.<\/p>\n
\nSeveral companies illustrated the capability of ChatGPT, and other large language models, to boost the efficacy of solutions in their sector.<\/p>\n