Mediaboss Marketing

The current wave of AI innovation is making smart TVs smarter.

At CES 2024 this week in Las Vegas, giant and startup TV manufactures alike are rolling out new AI integrations to power viewing, advertising and shopping.

Earlier this week, Telly a startup that gives people a free 4K TV in exchange for more ads debuted a new voice AI assistant called Hey Telly. Built with OpenAIs large language model, the chatbot helps operate the TV, chat with viewers and behave in ways similar to ChatGPT. With three possible characters to start and more to come the chatbot can also provide personalized recommendations based on whos watching. Its also exploring ways of using generative AI platforms like Midjourney and DALL-E to let people make generative AI images for TV screensavers.

Besides using OpenAIs LLMs, Telly also employs others for various tasks. For example, it uses a separate voice engine that can translate speech to text and then use that raw text for various AI models depending on uses.

We have brought in the models, all the processing, that we see fit for the use case, said Telly chief product officer Sascha Prueter.

Beyond startups, giants like Samsung, LG and Hisense are all adding AI to new TVs to improve picture and sound quality. For example, some use visual recognition to optimize the screen based on the content someone is watching or playing. Others are using AI to optimize audio by analyzing a rooms background noise and voices.

TV manufacturers also hope new AI features will make screens more useful beyond entertainment options include assisting people with workouts, providing telehealth tools, and becoming a hub for controlling other smart home devices. One end goal is to potentially replace other smart screens from companies like Amazon and Google by making the TV a central hub for everything in the home.

Some marketers who attended CES said they think new transparent TVs from LG and Samsung also powered by new AI processors offer compelling uses beyond showing viewers whats on the wall behind their big screen. For example, the newly unveiled devices which can operate like a normal TV or turn off to be clear as glass might someday be used by retailers to give passersby new ways to window shop.

AI is also enabling more options for commerce. Telly has added new AI-powered image recognition tools that show products to buy depending on whats on the screen. In a demo for Digiday during CES, a televised basketball game showcased products worn by players that can then be bought directly through Telly. Advertisers will also be able to sponsor product ads to show alongside other unpaid personalized recommendations.

Telly isnt the only TV startup thinking about AI and e-commerce. Another on the showroom floor at CES was Displace, which initially debuted a wireless TV at CES 2023. This year, the startups added a way to use hand gestures which the TV recognizes with AI to tell the TV to pause the show, analyze the screen for various products before Displace suggests similar products to buy.

Unlike Telly, Displace doesnt plan to run sponsored product ads, but one of its TVs does have an NFC reader so people can tap their preferred method of payment to buy something on Amazon or another platform. (Displace is also using OpenAIs technology along with various other sources.) Displaces new features also hope to ultimately provide recipes and show where to buy ingredients based on a cooking show someone is watching in their kitchen.

We are building a TV platform, said Displace founder and CEO Balaji Krishnan. This is exactly what happened with smartphones. Before smartphones, everything was SMS-based messaging. And even payments PayPal started with email, but now everything is integrated into the phone because companies like Apple and Google created a platform We are trying to create a contextual thing for the TV set.

Of course, the AI used in new cameras and voice tools also creates new privacy concerns: How do they make sure their faces, conversations, credit card data and other information arent leaked by one of the LLMs that process them? Prueter also said Tellys survey data is highly confidential and wont be fed back into ChatGPT or other public AI models.

To offer a sense of privacy, Tellys camera has a shutter in front and requires permission before taking a photo, but other sensors know when people are watching and how many are in the room. On Displaces devices, the camera pops up and can get closed into the screen whenever they dont want to use it.

Although ads are part of Tellys pitch people need to fill out a lengthy survey with more than 100 questions that helps to personalize features and advertising thats not the case for Displace. When asked about whether Displace plans to offer retargeting for advertisers, Krishnan said the company will never sell data because ads arent our thing.

What we are enabling is the advertisers now can actually pay their ads in a way that is interactive and transactional, Krishnan said. Because theyre already playing ads.Like if you look at the Super Bowl ads, theyre millions of dollars worth of ads, but theyre not interactive.

More:

At CES, new AI tools for TVs offer new features for both content and commerce - Digiday

Columbia engineers have built a new AI that shatters a long-held belief in forensicsthat fingerprints from different fingers of the same person are unique. It turns out they are similar, only weve been comparing fingerprints the wrong way!

Jan 10 2024 | By Holly Evarts | Photo Credit: Marco-Marcil Montoto, Columbia Engineering, generated with Dall-E

AI discovers a new way to compare fingerprints that seem different, but actually belong to different fingers of the same person. In contrast with traditional forensics, this AI relies mostly on the curvature of the swirls at the center of the fingerprint, as shown by the heatmap. Credit: Gabe Guo, Columbia Engineering; Midjourney generated silhouette.

From Law and Order to CSI, not to mention real life, investigators have used fingerprints as the gold standard for linking criminals to a crime. But if a perpetrator leaves prints from different fingers in two different crime scenes, these scenes are very difficult to link, and the trace can go cold.

Its a well-accepted fact in the forensics community that fingerprints of different fingers of the same person--intra-person fingerprints--are unique, and therefore unmatchable.

A team led by Columbia Engineering undergraduate senior Gabe Guo challenged this widely held presumption. Guo, who had no prior knowledge of forensics, found a public U.S. government database of some 60,000 fingerprints and fed them in pairs into an artificial intelligence-based system known as a deep contrastive network. Sometimes the pairs belonged to the same person (but different fingers), and sometimes they belonged to different people.

Credit: Gabe Guo and Aniv Ray/Columbia Engineering

Over time, the AI system, which the team designed by modifying a state-of-the-art framework, got better at telling when seemingly unique fingerprints belonged to the same person and when they didnt. The accuracy for a single pair reached 77%. When multiple pairs were presented, the accuracy shot significantly higher, potentially increasing current forensic efficiency by more than tenfold. The project, a collaboration between Hod Lipsons Creative Machines lab at Columbia Engineering and Wenyao Xus Embedded Sensors and Computing lab at University at Buffalo, SUNY, was published today in Science Advances.

Once the team verified their results, they quickly sent the findings to a well-established forensics journal, only to receive a rejection a few months later. The anonymous expert reviewer and editor concluded that It is well known that every fingerprint is unique, and therefore it would not be possible to detect similarities even if the fingerprints came from the same person.

The team did not give up. They doubled down on the lead, fed their AI system even more data, and the system kept improving. Aware of the forensics community's skepticism, the team opted to submit their manuscript to a more general audience. The paper was rejected again, but Lipson, who is the James and Sally Scapa Professor of Innovation in the Department of Mechanical Engineering and co-director of the Makerspace Facility, appealed. I dont normally argue editorial decisions, but this finding was too important to ignore, he said. If this information tips the balance, then I imagine that cold cases could be revived, and even that innocent people could be acquitted.

While the systems accuracy is not sufficient to officially decide a case, it can help prioritize leads in ambiguous situations. After more back and forth, the paper was finally accepted for publication by Science Advances.

One of the sticking points was the following question: What alternative information was the AI actually using that has evaded decades of forensic analysis? After careful visualizations of the AI systems decision process, the team concluded that the AI was using a new kind of forensic marker.

The AI was not using minutiae, which are the branchings and endpoints in fingerprint ridges the patterns used in traditional fingerprint comparison, said Guo, who began the study as a first-year student at Columbia Engineering in 2021. Instead, it was using something else, related to the angles and curvatures of the swirls and loops in the center of the fingerprint.

Columbia Engineering senior Aniv Ray and PhD student Judah Goldfeder, who helped analyze the data, noted that their results are just the beginning. Just imagine how well this will perform once its trained on millions, instead of thousands of fingerprints, said Ray.

The team is aware of potential biases in the data. The authors present evidence that indicates that the AI performs similarly across genders and races, where samples were available. However, they note, more careful validation needs to be done using datasets with broader coverage if this technique is to be used in practice.

This discovery is an example of more surprising things to come from AI, notes Lipson. Many people think that AI cannot really make new discoveriesthat it just regurgitates knowledge, he said. But this research is an example of how even a fairly simple AI, given a fairly plain dataset that the research community has had lying around for years, can provide insights that have eluded experts for decades.

He added, Even more exciting is the fact that an undergraduate student, with no background in forensics whatsoever, can use AI to successfully challenge a widely held belief of an entire field. We are about to experience an explosion of AI-led scientific discovery by non-experts, and the expert community, including academia, needs to get ready.

Continue reading here:

AI Discovers That Not Every Fingerprint Is Unique - Columbia University

Listen to this story. Enjoy more audio and podcasts on iOS or Android.

Your browser does not support the

AS PUPILS AND students return to classrooms and lecture halls for the new year, it is striking to reflect on how little education has changed in recent decades. Laptops and interactive whiteboards hardly constitute disruption. Many parents bewildered by how their children shop or socialise would be unruffled by how they are taught. The sector remains a digital laggard: American schools and universities spend around 2% and 5% of their budgets, respectively, on technology, compared with 8% for the average American company. Techies have long coveted a bigger share of the $6trn the world spends each year on education.

When the pandemic forced schools and universities to shut down, the moment for a digital offensive seemed nigh. Students flocked to online learning platforms to plug gaps left by stilted Zoom classes. The market value of Chegg, a provider of online tutoring, jumped from $5bn at the start of 2020 to $12bn a year later. Byjus, an Indian peer, soared to a private valuation of $22bn in March 2022 as it snapped up other providers across the world. Global venture-capital investment in education-related startups jumped from $7bn in 2019 to $20bn in 2021, according to Crunchbase, a data provider.

Then, once covid was brought to heel, classes resumed much as before. By the end of 2022 Cheggs market value had slumped back to $3bn. Early last year investment firms including BlackRock and Prosus started marking down the value of their stakes in Byjus as its losses mounted. In hindsight we grew a bit too big a bit too fast, admits Divya Gokulnath, the companys co-founder.

If the pandemic couldnt overcome the education sectors resistance to digital disruption, can artificial intelligence? ChatGPT-like generative AI, which can converse cleverly on a wide variety of subjects, certainly looks the part. So much so that educationalists began to panic that students would use it to cheat on essays and homework. In January 2023 New York City banned ChatGPT from public schools. Increasingly, however, it is generating excitement as a means to provide personalised tutoring to students and speed up tedious tasks such as marking. By May New York had let the bot back into classrooms.

Learners, for their part, are embracing the technology. Two-fifths of undergraduates surveyed last year by Chegg reported using an AI chatbot to help them with their studies, with half of those using it daily. Indeed, the technologys popularity has raised awkward questions for companies like Chegg, whose share price plunged last May after Dan Rosensweig, its chief executive, told investors it was losing customers to ChatGPT. Yet there are good reasons to believe that education specialists who harness AI will eventually prevail over generalists such as OpenAI, the maker of ChatGPT, and other tech firms eyeing the education business.

For one, AI chatbots have a bad habit of spouting nonsense, an unhelpful trait in an educational context. Students want content from trusted providers, argues Kate Edwards, chief pedagogist at Pearson, a textbook publisher. The company has not allowed ChatGPT and other AIs to ingest its material, but has instead used the content to train its own models, which it is embedding into its suite of learning apps. Rivals including McGraw Hill are taking a similar approach. Chegg has likewise developed its own AI bot that it has trained on its ample dataset of questions and answers.

What is more, as Cheggs Mr Rosensweig argues, teaching is not merely about giving students an answer, but about presenting it in a way that helps them learn. Understanding pedagogy thus gives education specialists an edge. Pearson has designed its AI tools to engage students by breaking complex topics down, testing their understanding and providing quick feedback, says Ms Edwards. Byjus is incorporating forgetting curves for students into the design of its AI tutoring tools, refreshing their memories at personalised intervals. Chatbots must also be tailored to different age groups, to avoid either bamboozling or infantilising students.

Specialists that have already forged relationships with risk-averse educational institutions will have the added advantage of being able to embed AI into otherwise familiar products. Anthology, a maker of education software, has incorporated generative-AI features into its Blackboard Learn program to help teachers speedily create course outlines, rubrics and tests. Established suppliers are also better placed to instruct teachers on how to make use of AIs capabilities.

Bringing AI to education will not be easy. Although teachers have endured a covid-induced crash course in education technology, many are still behind the learning curve. Less than a fifth of British educators surveyed by Pearson last year reported receiving training on digital learning tools. Tight budgets at many institutions will make selling new technology an uphill battle. AI sceptics will have to be won over, and new AI-powered tools may be needed to catch AI-powered cheating. Thorny questions will inevitably arise as to what all this means for the jobs of teachers: their attention may need to shift towards motivating students and instructing them on how to best work with AI tools. We owe the industry answers on how to harness this technology, declares Bruce Dahlgren, boss of Anthology.

If those answers can be provided, it is not just companies like Mr Dahlgrens that stand to benefit. An influential paper from 1984 by Benjamin Bloom, an educational psychologist, found that one-to-one tutoring both improved the average academic performance of students and reduced the variance between them. AI could at last make individual tutors viable for the many. With the learning of students, especially those from poorer households, set back by the upheaval of the pandemic, such a development would certainly deserve top marks.

Read more from Schumpeter, our columnist on global business:Meet the shrewdest operators in todays oil markets (Jan 3rd)Can anyone bar Europe do luxury? (Dec 20th)Boneheaded anti-immigration politicians are throttling globalisation (Dec 14th)

Also: If you want to write directly to Schumpeter, email him at [emailprotected]. And here is an explanation of how the Schumpeter column got its name.

Read more here:

AI can transform education for the better - The Economist

CIOs are under increasing pressure to deliver AI across their enterprises a new reality that, despite the hype, requirespragmatic approaches to testing, deploying, and managing the technologies responsibly to help their organizations work faster and smarter.

The top brass is paying close attention. Seventy-one percent of business leaders expect AI and ML to have a worldwide impact, according to the WorkdayC-Suite Global AI Indicator Report.Business leaders are excited about what AI and ML could do for their organizationsespecially operational efficiency, better decision-making, and competitive advantage, says the report.

That excitement is creating an acute sense of urgency among IT leaders and their teams. AmongIT leader respondentsin the AI Indicator study, the No. 1 concern is that IT leaders will face pressure to make difficult decisions about where to apply AI and ML.

Those decisions will have far-reaching impact across the organization. IT leaders expect AI and ML to drive a host of benefits, led by increased productivity, improved collaboration, increased revenue and profits, and talent development and upskilling. As AI and ML evolve, so will the skills of the humans supporting these initiatives.

A lot of new roles are going to emerge in the next couple of years as some of the existing roles become less important, says Prashant Nema, global CIO at Arch Capital Services, in the report. There has to be an ongoing focus on making sure that your talent is continuously learning and developing.

A data-driven foundation

Of course, a dose of caution is in order, particularly with newer AI offshoots such as generative AI. IT leaders understand that the models are only as good as the information on which they are educated. Outrageously inaccurate ChatGPT musings are just an opener for what could later be catastrophic mistakes predicated on bad data.

In the AI Indicator Report, almost 60% of IT leaders conceded that their companys data is somewhat or completely siloed, making it difficult for AI and ML to leverage fully.

Our biggest blocker to unleashing the power of AI is uncertainty over the integrity of the dataset its working from, Dan Cohen, CIO and director of operations at The Amenity Collective, says in the report. Our internal data and adherence to process is where our focus is, and we dont necessarily want to leap ahead until we feel like we have a stable footing there.

Ensuring data integrity is part of a broader governance approach organizations will require to deploy and manage AI responsibly.TheNISTs AI Risk Management Frameworkis a good place to start, providing IT teams with guidance on the design, development, and use of responsible AI products and services

Despite the risks, CIOs understand that embracing AI is a question of when and how, not if. Organizations across all industries are moving forward with pilots and production.

AI and ML are a game-changer for business, Chandler Morse, vice president of corporate affairs at Workday, says in a recentpodcast. The thing thats dawning on everyone is that its tough to see any sector in the economy that isnt going to be adopting these tools.

In the same podcast, Ajay Agrawal, professor at the University of Torontos Rotman School of Management, recommends that every company pick at least one AI project in a key business area to get started: Unlike any other tools our human civilization has used before, AI learns, so it gets better with use. The people who sit on the sidelines will miss all that learning time, and those that are building their AI now [will gain] the advantage.

For more insights, strategies, and best practices for IT leaders, visitWorkdays CIO Insights.

Go here to see the original:

CIO insights: Whats next for AI in the enterprise? - CIO

Canaccord Genuity is laying out some names poised to capitalize on the new stage of AI set to unfold in 2024. "Enthusiasm for AI's potential to improve human life and create business value has captivated investors and played a meaningful role in recent market optimism," wrote analyst Susan Anderson. "As we enter 2024, we anticipate a marked shift from businesses creating AI strategies towards executing them, with a focus on measuring and delivering ROI." Makers of large language models such as Alphabet and Microsoft, and chipmakers such as Nvidia dominated artificial intelligence investing in 2024, boosting the tech-heavy Nasdaq Composite. While those stocks may have won 2023, Anderson believes the market is only in the "early innings of a Cambrian explosion of AI-native applications." Here are some of the companies across sectors best situated to take advantage of the next stage of AI growth: Canccord Genuity sees a plethora of opportunities in the technology space beyond the pure play behemoths that prospered in 2024. Those beneficiaries include software vendors such as HubSpot and those providing the infrastructure for data quality, scalability and model running such as MongoDB . The firm also views cybersecurity stocks as well positioned to benefit as AI attacks increase in sophistication and defenders and cybercriminals battle it out with an array of new generative AI tools. "We believe security vendors will address AI to a much greater extent in 2024 and beyond to stay ahead of the curve or risk losing ground due to competition," analysts wrote. "In the software security section of the report, we address how our coverage companies are implementing and/or enhancing AI strategies." But technology isn't the only industry poised to benefit from AI tailwinds this year and beyond. In fact, the firm labeled a slew of beneficiaries within the health and medical field such as Dexcom and Illumina . DXCM 1Y mountain Dexcom shares over the last year AI implementation can be harnessed to monitor issues such as diabetes can help users monitor glucose levels and assist in DNA sequencing by predicting disease-causing mutations, the firm wrote. Even agriculture is poised to see a significant impact from artificial intelligence, with Deere taking a leadership role in the industry. "Through its centrality within farming operations and vast array of data-collecting sensors, DE's equipment increasingly provides the foundation for many of the AI efforts presently underway in the industry," the firm wrote. "Recent acquisitions focused on precision ag and AI underscore the point." DoubleVerify , Mobileye , Doximity and Amazon were also included in the list. CNBC's Michael Bloom contributed reporting

Go here to see the original:

How to play the next stage of AI in 2024, according to Canaccord Genuity - CNBC