AI Salesman: China's e-commerce platforms set to get AI-hosts for announcing deals during sale livestreams

China’s massive live-streaming e-commerce sector is known for peddling everything from lipsticks, food and drink to smartphones, cars and even a rocket launch service. It can easily generate tens of billions of dollars in sales overnight during major retail events like Singles’ Day.

While major Chinese cities like Shenzhen and Hangzhou are in a race to establish a global live-streaming e-commerce hub, a number of Chinese tech companies are now pushing generative artificial intelligence (AI) technology to create virtual live-streaming hosts capable of round-the-clock sessions, further raising the stakes in this sector.

Generative AI refers to the algorithms, such as those that power ChatGPT and similar services, which can be used to create new content, including audio, code, images, text, simulations and videos.

This application for generative AI underscores how the technology is bringing sweeping disruptions to certain traditional occupations in China and the broader job market.

The emergence of new generative AI-powered virtual hosts has the potential to significantly disrupt the livelihoods of over 400,000 human live-streamers across various platforms, including Taobao Live from Alibaba Group Holding, WeChat from Tencent Holdings, Douyin and Kuaishou Technology, which is owned by ByteDance. Even Alibaba, which owns the South China Morning Post, is part of this landscape.

According to Hugo Huang, the founder of a Guangzhou-based virtual host provider called Sansongshuzi, the cost of employing a virtual host for online marketing campaigns is much lower compared to hiring a human live-streamer. He states that his company offers a virtual host for only 500 yuan (US$70) per month, capable of appearing in short videos or live-streaming sessions.

In contrast, the typical expenses involved in hiring a human host and renting a physical studio would amount to about 50,000 yuan per month, making the use of virtual hosts a far more cost-effective option for businesses, both small vendors and major brands alike. This significant cost difference is likely to accelerate the adoption of AI-powered virtual hosts and may lead to the displacement of many human live-streamers from their current jobs.

Several major brands in China’s live-streaming e-commerce sector have shown keen interest in exploring the possibilities of generative AI-driven virtual hosts. According to Gao Zilong, the COO of WH Zones, a Beijing-based start-up specializing in virtual influencers, large consumer brands, including a local electronics giant and an international cosmetics conglomerate, are collaborating with his company to launch online campaigns using virtual live-streamers. These virtual hosts are designed to resemble specific human celebrities.

Gao believes that combining the technology of virtual hosts with the influence of popular human celebrities within their local fan base can significantly enhance a brand’s online sales.

Currently, there are two types of virtual hosts being developed: two-dimensional (2D) and three-dimensional (3D) simulation technologies. While 2D hosts are more cost-effective, 3D hosts offer greater creative potential.

However, Gao points out that in order for 3D virtual hosts to match the capabilities of human live-streamers, the overall cost of using them would be even higher than hiring a live host.

Despite the enthusiasm surrounding virtual hosts in China’s live-streaming e-commerce market, this AI application is still in its early stages. Zhang Yi, the CEO of research firm iiMedia, comments that the application is currently limited, and there is ongoing debate about whether this business model can be widely adopted.

Apart from the cost benefits, Zhang believes that young people online are likely to be attracted to digital personas and are more open to embracing this new technology. The appeal of virtual hosts to the younger generation could contribute to the further development and adoption of this AI-driven approach in the live-streaming e-commerce market.

During a recent corporate event organized by Baidu in Hangzhou, China, the renowned online search giant and advocate for artificial intelligence (AI) showcased a new workflow based on its large language model, which is also the technology behind its chatbot Ernie Bot. This workflow allows users to input specific keywords related to products, sales targets, and content length, and in return, it generates a script that human hosts can use to introduce products during live-streaming sessions. Additionally, the AI-powered solution provides other content that hosts can use to engage with their online audience effectively.

Baidu shared a success story of a vendor who used the same AI solution to sell tickets to Shanghai Disneyland, resulting in impressive sales of 540,000 yuan in just one month.

While AI brings significant advantages in generating content for live hosts and creating virtual live-streamers, some platforms still prefer human performance. Huang from Sansongshuzi mentioned that platforms like Douyin and Kuaishou continue to favor human live-streamers over virtual ones. He emphasized that virtual live-streamers might not receive as much attention on these platforms, but e-commerce platforms like Taobao, owned by Alibaba, have not limited the visibility of virtual live-streamers on their platforms. The competition between AI-powered virtual hosts and human live-streamers is still ongoing, and their coexistence is evident across different platforms in the live-streaming e-commerce market.

from Firstpost Tech Latest News https://ift.tt/0OqjTJE

It's Official: Elon Musk completes Twitter's rebranding to X, tweets will be called ‘sending an X’

Elon Musk, the billionaire CTO of Twitter, recently shared some ambitious plans to significantly transform the social media platform, mainly replacing the well-known blue bird logo with a new identity, simply named ‘X.’ Musk tweeted that “X dot com now points to Twitter dot com” and mentioned that the interim X logo would go live later that day. Subsequently, he revealed the temporary logo for the platform.

Following the announcement, Twitter has begun the gradual process of adopting the new X branding. The official Twitter handle has already been changed to X, along with a new X profile picture. Elon Musk himself has also updated his profile picture on the platform.

As of writing this article, Musk has already updated the Twitter interface and replaced the blue bird logo with the new X logo. In fact, Musk has also changed Twitter’s Profile name to X.

Despite the usual enthusiasm for changes, this particular move has received mixed reactions. Some users took to the microblogging site to express their concerns, labelling it a ‘bad idea.’

CEO Linda Yaccarino, also tweeted out the unveiling of the new logo, while tweeting, “X is here! Let’s do this.”

X is here! Let’s do this. pic.twitter.com/1VqEPlLchj

— Linda Yaccarino (@lindayacc) July 24, 2023

Musk also plans to change the terminology that people using Twitter have developed over the years. For example, tweets from now on will be called “Xs”. Basically, “sending a tweet,” will now become “sending an X.”

Elon Musk says tweets will now be called ‘an X’ pic.twitter.com/ikXRa1ERDb

— Culture Crave (@CultureCrave) July 23, 2023

Of all the things that have gone with the rebranding exercise, changing Tweets to an X is one of the riskiest things Musk could have done, as explained succinctly by a user

The dumbest thing about this is how every other social platform WISHES they created a unique and renowned verb like “Tweet” or “Retweet” – that kind of brand recognition is near impossible to replicate intentionally pic.twitter.com/zSeAFEHlgV

— Del Walker (@TheCartelDel) July 23, 2023

People, especially Indian Twitter users had some hilarious takes on this:

Please don’t upload videos on twitter anymore, it will be known as X Videos

— Roshan Rai (@RoshanKrRaii) July 24, 2023

Tweet – Retweet
X – ReX ??

— nobu (@Senaapatii) July 24, 2023

from Firstpost Tech Latest News https://ift.tt/sI3FXDw

ChatGPT’s Sam Altman creates new cryptocurrency called Worldcoin, meant ‘only for humans’

Worldcoin, a cryptocurrency initiative founded by OpenAI’s CEO Sam Altman, is commencing its launch on Monday.

At the core of this project lies the World ID, an exclusive account accessible only to real individuals. To obtain a World ID, customers must undergo an in-person iris scan using Worldcoin’s ‘orb,’ a silver ball resembling the size of a bowling ball. Once the orb confirms the person’s authenticity, a World ID is created.

Tools for Humanity, a company based in San Francisco and Berlin, is the driving force behind Sam Altman’s latest crypto project.

During its beta phase, the project amassed 2 million users. Now, with the official launch on Monday, Worldcoin is expanding its ‘orbing’ operations to 35 cities across 20 countries. As an incentive, individuals signing up in select countries will receive Worldcoin’s cryptocurrency token, WLD.

The incorporation of cryptocurrency into World IDs is crucial as it enables the storage of these IDs on blockchain technology, ensuring privacy and immunity from control by any single entity.

World IDs are deemed essential in the era of advanced AI chatbots like ChatGPT, which produce remarkably human-like language. These IDs can be employed to distinguish between genuine individuals and AI bots online.

Binance, the largest cryptocurrency exchange, has announced its intention to list Worldcoin, with trading expected to tentatively begin on Monday at 09:00 GMT.

Altman also shared how Worldcoin can contribute to reshaping the economy under the influence of generative AI. He envisions AI empowering people and foresees significant economic implications. Altman highlights universal basic income (UBI) as an example—a social benefits program where each individual receives regular payments. With AI taking over more human tasks, Altman believes UBI can combat income inequality. Moreover, since only real people can possess World IDs, it could help reduce fraud when implementing UBI.

Although Altman acknowledges that a world with UBI is distant and lacks a clear entity to oversee it, he sees Worldcoin as a stepping stone towards making it a reality. He emphasizes the importance of experimenting with innovative ideas to determine the best course of action.

from Firstpost Tech Latest News https://ift.tt/lmF84pQ

Christopher Nolan insists people watch 'Oppenheimer' in IMAX and here's why they should

Oppenheimer will probably go down as the best film of 2023. The Christopher Nolan directorial certainly has been the most anticipated film of the year. Nolan, the filmmaking genius that he is has used some of the highest-resolution film cameras available, specifically IMAX 65mm and Panavision 65mm, to shoot the movie. This is unusual because IMAX cameras are typically used for fast-paced action shots, while more convenient cameras are used for other scenes in movies.

Nolan has often recommended watching “Oppenheimer” in true IMAX format. Although the film has been digitized to be compatible with all screens, it is best experienced in specific IMAX theatres. However, only a limited number of theatres worldwide are showing the movie in its most epic and immersive form, in true IMAX. In fact only 19 theatres in the US and just over 30 theatres the world over are showing the film in true IMAX. In case you’re one of the lucky ones who live close to such a theatre, here’s why you should watch the film in only True IMAX.

The aspect ratio gives more room and more depth
IMAX movie screens are well known for their instantly recognizable tall aspect ratio, usually 1.43:1 or 1.90:1. This aspect ratio provides a larger and more immersive visual experience compared to standard movie screens. The taller screen allows for a more encompassing view of the film, making the audience feel like they are right in the middle of the action.

The large screen fills the audience’s peripheral vision, creating a sensation of movement when the camera pans or moves. It offers a truly cinematic and captivating experience for the viewers.

The seating in True IMAX theatres
In addition to the unique aspect ratio, IMAX theatres also have steeper stadium seating arrangements within the cinema hall. This seating design ensures that each viewer has an unobstructed view of the screen, enhancing the overall movie-watching experience.

The combination of the tall aspect ratio and specialized stadium seating in IMAX theatres not only delivers an exceptional visual experience but also leads to better audio quality. The immersive setup allows the audience to feel fully engaged with the movie’s action and sound, creating a more captivating and memorable cinematic journey.

The glory of 16K
IMAX movies are shot using special cameras that can record with a larger frame, typically three times the theoretical horizontal resolution of a conventional 35mm film. Typically, IMAX uses a proprietary version of the 65mm film format, which has a native resolution of 16K. That’s 16 times the resolution of a 4K panel.

Needless to say the image is super sharp, with great depth. After the filming is complete, IMAX further enhances the footage through a proprietary image processing technique. This image enhancement is applied to every frame in the movie, ensuring that the final result delivers the clearest and sharpest picture quality possible. The goal is to present the movie exactly as the director intended, providing the audience with an immersive and true-to-life cinematic experience.

With the combination of larger frame recording, high-resolution cameras, and sophisticated image processing, True IMAX movies offer unparalleled visual quality, making them stand out as a premium and awe-inspiring movie experience.

from Firstpost Tech Latest News https://ift.tt/kpR3ZDK

GoI rejects Chinese EV maker, BYD's plan to set up a $1 billion dollar plant in India; Here’s why

A few days ago, BYD, also known as Build Your Dreams, one of the world’s largest electric vehicle (EV) manufacturers, announced that it plans to invest $1 billion in India and set up a factory for making EVs and batteries in the country. They also announced that the plant will be set up in partnership with Hyderabad-based Megha Engineering and Infrastructures Ltd.

However, recent developments show that the Indian government has rejected BYD’s proposal. The companies had jointly applied to the Department for Promotion of Industry and Internal Trade (DPIIT) to set up an electric vehicle plant in Hyderabad.

Why India rejected a $1 billion plant from BYD
During the evaluation process, various departments were consulted for their input, and concerns related to security, particularly regarding Chinese investments in India, were raised.

According to officials familiar with the discussions, existing regulations do not permit such investments from China. The proposal put forth by the partnership aimed to manufacture 10,000-15,000 electric cars annually, with Megha providing the capital and BYD contributing the technology and know-how.

This rejection is consistent with the ongoing tensions between India and China, which have been affecting the business endeavours of Chinese companies in India. Notably, Great Wall Motor, another Chinese automobile manufacturer, faced a similar setback last year when its proposed $1 billion investment plan was unable to secure government approvals.

The background to these investment policies lies in India’s 2020 policy, which was implemented after hostilities with China. The policy imposed restrictions on Foreign Direct Investments originating from China, mandating prior approval for investments from nations sharing land borders with India. This policy has been perceived by experts as anti-China. In May of this year, the Indian government also introduced measures to scrutinize Chinese companies attempting to bypass this rule by using shell companies and availing subsidies usually denied to Chinese businesses.

BYD’s growing presence in India
Despite this setback, BYD currently operates in the Indian market, offering two fully-electric cars – Atto 3 and e6. The company plans to launch its Seal EV in India by the end of the year, which was showcased at the 2023 Auto Expo. The introduction of the Seal EV is expected to follow a similar route as the Atto 3 and e6, using semi-knocked down kits (SKD).

Industry insiders have told Firstpost that there is a substantial fear that Chinese EV brands manufacturing in India will bring several challenges. First, Chinese companies operating in India have a history of fraudulently dodging taxes and underreporting their tax liabilities. Second, bringing physical components from China to a Chinese-owned company in India opens up a Pandora’s box of potential issues and problems.

India’s Tesla Dreams
Meanwhile, discussions are ongoing between the Indian government and Tesla, owned by US billionaire Elon Musk, for establishing a factory in India. While the government has shown interest in Musk’s proposals, it has not offered any additional incentives or tax breaks, requiring Tesla to apply under the existing incentive plans for EVs.

from Firstpost Tech Latest News https://ift.tt/nKVSZUk

X Marks The Spot: Elon Musk starts Twitter's makeover into a ‘super app’ with many new upcoming features

Ever since news broke that Elon Musk would be taking over Twitter, industry experts, especially those who have been following Elon Musk and his various business ventures were saying that Musk will one day turn Twitter into a super app.

Basically, Twitter will be made into an app that does pretty much everything, from being social media platform, to a messaging app, to a payments and banking app and, even to let users order food, book a cab, or just shop online. With Twitter now being rebranded as X, and Musk stating that he will be ‘killing all birds’ The transition everyone knew was on the way, is now finally happening.

pic.twitter.com/IwcbqMnQtA

— Elon Musk (@elonmusk) July 23, 2023

Elon Musk made an announcement on Sunday stating that “X dot com” now directs to Twitter dot com. He also tweeted that the interim X logo would be live later in the day. Mr. Musk has been vocal about his aspiration to create an all-in-one app that combines the functionalities of social media and payment apps, similar to China’s WeChat, which is owned by the Chinese tech giant Tencent.

WeChat, a crucial part of Chinese society, boasts more than 1 billion users who rely on it for a wide array of tasks, including social media, messaging, food ordering, and cab payments. Its significance is such that it is often referred to as “the app for everything.”

In the past, Mr. Musk openly expressed that his acquisition of Twitter, which stirred controversy, served as a catalyst for his vision of creating “X,” an all-encompassing app he envisions as the “everything app.”

Twitter’s CEO Linda Yaccarino, shed some light on what new features will come with X, the everything app, saying that X’s unlimited interactivity is powered by AI, and “it will connect us all in ways we’re just beginning to imagine.”

X is the future state of unlimited interactivity – centered in audio, video, messaging, payments/banking – creating a global marketplace for ideas, goods, services, and opportunities. Powered by AI, X will connect us all in ways we’re just beginning to imagine.

— Linda Yaccarino (@lindayacc) July 23, 2023

There’s absolutely no limit to this transformation. X will be the platform that can deliver, well….everything. @elonmusk and I are looking forward to working with our teams and every single one of our partners to bring X to the world.

— Linda Yaccarino (@lindayacc) July 23, 2023

From Linda’s Tweets above and the rumours of the features that some of Twitter’s top engineers have been working on, here are just some of the features that Elon Musk’s X super app will have

Centred around audio and video content
While Twitter is very text heavy, it will transition into audio, video and photographic content more as they are easier to interact with. Having said that, Twitter will still continue to support text but will urge creators to supplement those textual content with video or images.

Integration with food ordering apps, cab booking apps
Because of Musk’s fascination with WeChat, we can assume that some of the most popular features of the Chinese super app will be added to Musk’s X, the everything app. Some of the most common features would be social messaging, ordering food, and booking train/flight tickets or cabs.

E-commerce platform or an online marketplace
Almost every social media app has dabbled with the idea of having its own e-commerce platform, or at the very least, a marketplace. And although the way the idea may have been implemented differently by different social media platforms, Twitter’s online shopping platform is pretty rudimentary. That is all set to change because of X. It will be interesting to see if Musk’s X platform becomes a threat to Jeff Bezos’ Amazon in any way

A UPI-like payment and banking system
Payments and banking will be a cornerstone for X. Musk has indicated several times that Twitter, now X, needs its own payment system and that he hopes to start an online bank within the Twitter ecosystem. Given how many of X’s proposed features are based on buying and selling stuff services, it only makes sense to have a banking system within the app itself.

A dating platform
Twitter has already announced that it will now allow verified companies to post vacancies they have at their companies, muck like LinkedIn. There is a very possibility that Musk and Linda may decide to add a dating feature to the app. As anyone who has spent any time on the Internet would know, almost all social sites can be made into a dating site.

Moreover, Musk has at times encouraged people on Twitter to find their partners using the social media platform.

from Firstpost Tech Latest News https://ift.tt/Kk96R0C

Why Brits may not be able to use Apple's iMessage, FaceTime anymore?

Apple has announced that it will remove services like FaceTime and iMessage from the UK rather than compromise on security if new proposals in the Investigatory Powers Act (IPA) 2016 become law.

The UK government is seeking to update the IPA and wants messaging services to get approval from the Home Office for their security features before releasing them to the public.

A little too much oversight
Currently, the Home Office can demand security features to be disabled without informing the public, but the proposed update would require immediate action without a review or independent oversight process.

Messaging services, including WhatsApp and Signal, have expressed opposition to a clause in the Online Safety Bill that would allow the communications regulator to mandate the installation of technology to scan for child-abuse material in encrypted messaging apps and other services. They have refused to comply with this provision, with Signal even threatening to leave the UK.

Apple has also voiced its objections to the plan, standing in solidarity with other messaging platforms in prioritizing user privacy and security.

The government has launched an eight-week consultation on the proposed amendments to the IPA, which already allows the storage of internet browsing records for 12 months and permits the bulk collection of personal data. The government claims that the amendments are not about granting new powers but rather updating the act to align with current technology.

Why the Investigatory Powers Act is problematic
Apple has consistently opposed the Investigatory Powers Act, which has been criticized as a “snooper’s charter” by its detractors. In response to the current consultation on proposed amendments to the act, Apple has submitted a nine-page document outlining its objections.

Some aspects of the amendment require tech companies to inform the Home Office about any changes to product security features before their release. Non-UK-based companies need to comply with changes that would impact their products globally, such as providing a backdoor to end-to-end encryption. Tech companies will also need to take immediate action if the Home Office issues a notice to disable or block a feature, without any consultation

Apple’s stance is based on these is issues is as follows:

1. It refuses to make security feature changes that could weaken products for all users to accommodate the demands of one country.
2. Some changes may necessitate issuing a software update, making them impossible to implement secretly.
3. Apple views the proposed amendments as a serious and direct threat to data security and information privacy, not only for individuals in the UK but also for people outside the country.

The company firmly advocates for maintaining strong data security and privacy standards and expresses concerns that complying with the proposed changes could have far-reaching consequences.

Pulling up the national security card
In response to the concerns raised, the Home Office stated that the Investigatory Powers Act’s primary goal is to safeguard the public from criminals, child sex abusers, and terrorists. The government emphasizes that they continuously review all legislation to ensure it remains robust, and the current consultation is part of that ongoing process. No final decisions have been made yet.

The issue highlights a clash between the government’s aim to strengthen national security and protect the public and the concerns raised by technology companies and experts about potential implications for data security and privacy. The outcome of the consultation and the subsequent decisions will likely have significant implications for the relationship between tech companies and government surveillance policies in the UK.

from Firstpost Tech Latest News https://ift.tt/IjE23fS

JournoGPT: Google is secretly showcasing its generative AI for journalists to major US newspapers

Google has secretly unveiled an AI tool known as Genesis, capable of producing news stories for major newspapers like The New York Times, The Washington Post, and The Wall Street Journal.

According to reports from The New York Times, this signals a potential influx of AI-generated or AI-assisted content on the internet.

Although Google denies intending to replace journalists, they assert that Genesis will act as a “personal assistant” for journalists, automating certain tasks to allow more time for others.

Google downplaying its AI’s capability to get Journalists on board
Media executives who have seen a demo of the AI chatbot revealed that the tool can not only write and edit stories but give them headlines, straplines and summaries, all without the input of a trained media professional.

Media executives expressed concerns about Google’s pitch, feeling it downplayed the effort and skill required to create accurate and well-crafted news stories.

Publications, such as CNET, Gizmodo, and BuzzFeed, have already embraced AI-generated news stories, but they often suffered from errors and plagiarism issues.

The entrance of an influential company like Google into this space could accelerate the adoption of AI technology by media outlets, increasing pressure on traditional journalists.

“Goodness gracious,” tweeted The Information founder Jessica Lessin. “Let it be said, journalists don’t need Google to write their articles as ‘a personal assistant.’ And anything that Google (or any AI) could write has no real original reporting value.”

Google is killing the news
Google maintains that its tools are not meant to replace journalists entirely, but rather offer assistance with tasks like generating headlines or different writing styles.

“Quite simply, these tools are not intended to, and cannot, replace the essential role journalists have in reporting, creating and fact-checking their articles,” Google spokesperson Jenn Crider told the NYT. “For instance, AI-enabled tools could assist journalists with options for headlines or different writing styles,” she added.

Recently, Google unveiled a new AI-powered search interface called “Search Generative Experience” (SGE) that can summarize webpages through “AI snapshots,” potentially undermining human-based journalism further.

Hallucinations mean reputations are on the line
However, there are reasons to be sceptical about Google’s AI technology, as past instances like the AI chatbot Bard have shown its unreliability in distinguishing truth from fiction and generating misinformation at scale.

The media industry is exploring ways to incorporate generative AI into newsrooms, but concerns remain about potential job cuts and the ability of AI to produce coherent and more importantly fact-based truthful information.

Ultimately, the reputation of news publications is at stake, prompting media executives to carefully consider the implications before fully embracing Google’s AI offer.

from Firstpost Tech Latest News https://ift.tt/8byzF7p

Blind, leading the blind: Developers are using AI-generated data to train their AI bots

Although most AI models rely on data generated by humans, certain companies are now exploring the utilization of data produced by AI itself.

This concept, known as “synthetic data,” presents a promising opportunity for significant advancements in the AI ecosystem, although it also raises comparisons to an algorithmic ouroboros.

Feeding a data-hungry monster
According to the Financial Times, OpenAI, Microsoft, and the startup Cohere, valued at two billion dollars, are actively researching synthetic data to train their large language models (LLMs). The primary motivation behind this shift is the cost-effectiveness of synthetic data compared to expensive human-created data.

In addition to cost benefits, the issue of scale arises when training cutting-edge LLMs. The existing pool of human-generated data is already substantially utilized, and to further enhance these models, more data will likely be required.

According to Cohere’s CEO, Aiden Gomez, acquiring all the necessary data directly from the web would be ideal, but the reality is that the web is too chaotic and unstructured to represent the precise data needed. Therefore, companies like Cohere and others are already employing synthetic data to train their LLMs, although this approach is not widely publicized.

OpenAI’s CEO, Sam Altman, expressed confidence that synthetic data will eventually dominate, and Microsoft has started publishing studies on how it can enhance less sophisticated LLMs. Additionally, there are startups solely focused on selling synthetic data to other companies.

AI’s questionable integrity and reliability
However, critics point out a significant drawback: AI-generated data’s integrity and reliability might be questionable, as even AI models trained on human-generated data are known to make substantial factual errors. This process also carries the risk of creating messy feedback loops, labeled “irreversible defects” in a recent paper by Oxford and Cambridge researchers.

Nonetheless, companies like Cohere aim for a moonshot goal of self-teaching AIs that can generate their own synthetic data. The ultimate dream is to have models capable of asking their own questions, discovering new insights, and creating knowledge autonomously.

The problem with AI black box
Even developers who work on AI models have failed to understand how exactly do most AI algorithms work. Most AI studios are updating their existing AI models and LLMs by feeding them data, not by updating core code that controls the algorithm.

The AI block box is so opaque, that almost all AI models that have been allowed to operate freely have picked up some or other language on its own. Back in April, Google exec James Manyika admitted that even though they had not trained their experimental AI in Bengali, the model had picked up the language as well as few of its dialects, and perfected it.

This sort of behaviour, where an AI model teaches itself things is called emergent properties, and it is virtually impossible to stop AI models from doing it, without destroying it.

The way most AI models work is that it does not forget or rather erase anything that it has learnt. This includes things that are categorically wrong. Developers can put on filters on the output that it generates, but the AI model still has that piece of factoid within itself and uses it in its workings.

If developers are using faulty data or a data set that has been generated under a hallucination, the resulting AI bot will also generate faulty results.

And it’s not just that the results generated may be faulty, they can be biased as well. Several AI-generated content on Wikipedia, would be a great example of this. Biased articles were used to train a certain AI model, which in turn generated articles that were more biased than the previous one, to the point, where they were riddled with ‘facts’ that are hilariously incorrect.

from Firstpost Tech Latest News https://ift.tt/OwYZncF

Tesla, TSMC earnings predict grim future because of US-China tech war, investors worried

Tesla, TSMC and most tech manufacturing companies recently revealed their earning reports for the last quarter, and this has made investors a little jumpy. The latest earnings reveal that the tech sector is in for a really bad time, because of the US-China tech war

On Thursday, Swedish-Swiss multinational corporation ABB Ltd. issued a warning about slowing demand in China, one of its significant markets, reporting a 9 per cent decline in orders during the second quarter of the year. The company’s divisions in electrification, motion, and robotics all experienced reduced demand in the region. Similarly, TSMC and Tesla are in the same position.

ABB also faced decreased demand in Germany, indicating potential challenges in the broader industrial economy. ABB’s products, including motors, drives, controllers, and electrification solutions used in transport systems and factories, are often considered indicators of the overall health of the industrial sector.

These developments have caused concern among investors, who were hopeful that China’s decision to lift strict COVID-related restrictions at the end of the previous year would lead to an economic revival in the country. However, recent data showed that China’s economy grew at a fragile pace in the second quarter due to weakened demand both domestically and internationally, with post-COVID momentum diminishing rapidly.

In addition to ABB’s warning, Taiwanese chipmaker TSMC forecasted a 10 per cent drop in sales for 2023, attributing it to the global economic challenges that have affected the demand for chips used in various applications, including automobiles, cellphones, and servers.

Furthermore, several companies are grappling with the dilemma of maintaining profit margins after raising prices to counter the impact of soaring energy and raw material costs since the past year. As input costs are expected to ease, analysts caution that companies may be compelled to reduce prices to stay competitive, potentially affecting their bottom lines.

Tesla CEO Elon Musk also recently announced plans to cut prices on electric vehicles to shield against competition and uncertainties in the economic landscape, which had a negative impact on Tesla’s shares, declining nearly 4 per cent in pre-market U.S. trading. Overall, these developments are casting a shadow over the stock market, especially as the second-quarter earnings season unfolds. While European stocks regained some ground after early losses, S&P 500 and Nasdaq futures indicated lower openings.

“One day it seems like the world economy is falling apart, next day it’s fine. I don’t know what the hell is going on,” Musk told analysts on a conference call. “We’re in, I would call it, turbulent times.”

The recent corporate earnings results have set a pessimistic tone early in the earnings season. Consumer spending has been hindered by rising shopping and food expenses, along with high-interest rates, putting pressure on corporate profit margins.

Streaming video pioneer Netflix also contributed to the downbeat sentiment, as its second-quarter revenue fell short of analyst estimates, causing its shares to plummet by nearly 9 per cent in after-hours trading.

In the United States, earnings are predicted to fall by 8.2 per cent, which is a notable decline from the 0.2 per cent growth recorded in the first quarter and a significant reversal from the 8.4 per cent growth observed a year ago. Similarly, revenue is expected to decline by 0.8 per cent in the second quarter, down from the 13.6 per cent growth experienced a year ago.

Overall, these figures indicate a challenging economic environment, with a slowdown in corporate earnings and revenue growth compared to previous quarters and years. The ongoing impact of the pandemic, rising costs, and other economic factors are contributing to the subdued performance in the corporate sector.

from Firstpost Tech Latest News https://ift.tt/JQ4PZo9

Threads by Meta loses its top spot on App Store and Play Store to new viral AI photo app

Remini, an AI photo editing app, has dethroned Instagram Threads as the top app on the App Store, thanks to its popularity on TikTok. Initially launched in 2019, Remini introduced a generative AI feature last year, which TikTok users recently discovered can produce professional headshots by simply uploading their own selfies.

A TikTok user named @Gracesplace demonstrated how she used the app to create multiple impressive headshots with different outfits and poses. The video went viral, amassing over 42.9 million views. As more people followed suit, the demand for Remini skyrocketed, propelling it to the No. 1 position on the U.S. App Store on July 11th, surpassing Instagram Threads shortly after Meta’s new Twitter competitor launched. Over the last month, Remini has garnered more than 22 million downloads worldwide, with an estimated 20 million daily active users.

Another analytics firm, data.ai, also observed Remini’s rise, reporting that it overtook Threads on July 10th. In the U.S., the app recorded over 2 million downloads on iOS and Android in the past 10 days and generated $3.73 million in consumer spending from June 9-15, showing a remarkable 1,055% week-over-week increase.

To create the AI headshots, users need to select eight to 12 photos of themselves, including close-up selfies with various backgrounds, expressions, and angles, and then indicate their gender. Users then choose a “model” image that represents the desired style and composition for the generated photos. Remini offers several options under the “curriculum” subheading to produce professional shots, along with other styles like “trendy,” “travel,” “casual,” and even “Korean aesthetic.”

The AI takes a few minutes to generate the photos, which users can view by swiping left and right, similar to Tinder’s interface. This process is reminiscent of another AI photo app called Lensa AI, which also gained popularity on TikTok last year. However, Lensa AI faced criticism for using artists’ work as training data and enabling the creation of explicit content with the AI.

Like Lensa AI, Remini is a free download, but some of its features require a subscription, priced at $9.99 per week or $79.99 per year. Some users have reportedly used free trials to generate their AI images before cancelling the subscription.

Despite its popularity, Remini has received some backlash from users who were unhappy with the AI’s results. Many people noticed the app making unwelcome adjustments to their bodies in the generated photos, such as making them appear much skinnier or enhancing the size of their breasts. Additionally, there are concerns that the app’s AI-generated photos may deter people from hiring professional photographers, highlighting the broader implications of easily accessible AI technology.

from Firstpost Tech Latest News https://ift.tt/EFkL3hm