For executives, investors and employees in the entertainment and telecommunications industries, 2024 is set to disappoint.

Maybe that’s too grinchy. Some things will get better. The actors’ and writers’ strikes are over. The 2024 U.S. presidential election should help boost advertising dollars as global TV ad revenue is on pace to decline 18% this year, according to media investment firm GroupM.

Companies such as Warner Bros. Discovery and Disney cut thousands of jobs and dramatically slashed content costs to boost free cash flow and pay down debt. That could give investors a reason to be more sanguine about their business prospects next year. Disney recently restored its dividend for early 2024 after suspending it for more than three years.

Still, legacy media companies including Disney, Paramount Global, Warner Bros. Discovery and Comcast ’s NBCUniversal are trying to figure out what investors want since pulling back on a narrative of subscription streaming video growth that dominated 2020 and 2021. Warner Bros. Discovery and Comcast have outperformed the S&P 500 in 2023, though just barely. Disney and Paramount Global have underperformed.

The overriding narrative for 2024 appears to be one of uncertainty on three key fronts: interest rates, regulatory policy and overall growth prospects. The industry should have more clarity in 2025 on all three topics to propel it forward, said Corey Martin, managing partner at entertainment law firm Granderson Des Rochers. Next year will probably be defined by preparation for action rather than actual transformation, Martin said.

“2024 is probably going to be a year of sustained uncertainty,” said Martin. “It’s really a continuation of a pattern we’ve seen since the midpoint of 2022.”

Paypal's PYUSD is the first stablecoin issued by a "non-cryptocurrency" company. It represents a general shift in corporate attitudes towards stablecoins and a belief in the future of upcoming policy acceptance.

PayPal's stablecoin, PYUSD, is backed exclusively by US dollar deposits, such as short-term US Treasuries and equivalent assets. Managed by Paxos Trust Company, PYUSD can be converted to USD at a 1:1 ratio through PayPal or the Venmo app.

According to Paypal's press release:

PayPal customers who purchase PayPal USD will be able to:

Transfer PayPal USD between PayPal and compatible external wallets; Make inter-personal payments using PYUSD; Choose to make purchases with PayPal USD at checkout; Convert any PayPal-supported cryptocurrency to and from PayPal USD.

Why PayPal's announcement has a significant impact on cryptocurrency adoption

Until now, the only way to get access to payment stablecoins has been through cryptocurrency companies like Tether, Coinbase or Gemini. Now, with PayPal entering the market, millions of people will be able to enter the cryptocurrency space through one of the world's most widely used payment platforms.

One of the most underdeveloped parts of the cryptocurrency ecosystem is the actual access to the money," said Austin Campbell, former head of portfolio management at Paxos and partner at Zero Knowledge Consulting, on Leviathan News. From that perspective, it's hard to find a better option than PayPal. I think the biggest innovation here is adding a native stablecoin to the PayPal platform."

Campbell further stated that he believes "two and a half years of work went into this product before it was released." Rumors of PayPal's stablecoin development were reported back in 2021, when Jose Fernandez da Ponte, PayPal's vice president and general manager of blockchain, cryptocurrencies, and digital currencies, told the press, "It's still early days." Further rumors were leaked to the press confirming PayPal's plans, but in February of this year, the payments company announced that it was suspending its plans for a stablecoin product because Paxos was under investigation by the New York Department of Financial Services (NYDFS). Six months later, the company apparently decided that the regulatory environment was cool enough to launch their stablecoin.

PayPal chose Paxos to manage and issue their stablecoin, which means it will be fully stocked, have segregated funds, and publish regular transparency reports. Additionally, their stablecoin will be monitored by on-chain analytics firms such as Chainalysis and TRM to prevent illegal use. If criminal activity is involved, PayPal will be able to freeze the funds.

As a result of its relationship with Coin, Paxos drew the ire of regulators, with a series of claimed violations culminating in an order from the New York Department of Financial Services to halt the issuance of BUSD and the receipt of a Wells Notice from the SEC. The New York Department of Financial Services said the order was "due to several unresolved issues regarding the oversight of Paxos' relationship with CoinSafe".

"As directed by the New York Department of Financial Services, effective February 21, Paxos will cease issuing new BUSD tokens and will end its relationship with CoinSafe for the branded stablecoin BUSD," Paxos said in a statement.

Charles Cascarilla, CEO of Paxos, said, "The market has changed and the relationship with CoinSafe no longer aligns with our current strategic priorities."

Coin's relationship with Paxos allows them to mint BUSD directly from the exchange and transfer it to any blockchain of their choice.

The New York Department of Financial Services said, "The department has not authorized the use of CoinSecure-Peg BUSD on any blockchain, nor is CoinSecure-Peg BUSD issued by Paxos."

Now that PayPal is working with Paxos, it means that the investigation is over and the controversial issuer is now free to operate after a rigorous regulatory review.

PayPal Vs Meta.

While the announcement is still new, the reaction has been very different from that of Diem, the failed stablecoin developed by Meta. When Facebook first announced its entry into the market in 2021, the social network came under heavy fire from politicians, economists and activists in front of Congress.

At the time, Facebook was still reeling from the Cambridge Analytica scandal, which became a focal point of the 2020 elections. The company had not yet recovered its image and therefore reacted violently to the news of Diem.

Senator Elizabeth Warren expressed strong opposition to Facebook's "relaunch of cryptocurrencies and digital wallets". Additionally, she and Senators D-HI, D-OH, and others wrote in a letter that "Facebook is once again pursuing digital currency initiatives and has already launched a pilot of a payments infrastructure network, despite the fact that these programs are incompatible with the actual financial regulatory environment - not only for Diem specifically, but also for stablecoins. Diem specifically, but also for stablecoins in general."

According to Campbell, Facebook faces two problems.

First, unlike PayPal, Diem is an entirely new line of business for the company; Facebook is a social network used by more than 2 billion people worldwide, and also owns the Whatsapp and Instagram brands. Adding a payment service would instantly turn Facebook into a giant quasi-bank that could absorb all of its users overnight. Lawmakers and regulators are concerned that Facebook will further abuse its already considerable power to capitalize on the data it collects from users.

Not only does the social giant have access to your friends list, likes, posts, direct messages and geographic data, the addition of Diem has the potential to provide Facebook with unprecedented access to your personal finances. For a company whose image has already been tarnished by scandal, Diem is just a step too far.

The second problem is that Diem is more than just a dollar stabilizer; the deal also plans to issue a currency similar to the Special Drawing Rights (SDR), backed by a variety of foreign currency pairs such as the euro, yen, Australian dollar and Swiss franc. Economists are outraged by the proposal. In their view, global access to a basket of currencies would weaken the ability of central banks to control domestic monetary transmission. What poor schmuck would buy their worthless bonds if their citizens had easy global access to a less volatile stablecoin?

Diem has never gotten out of regulatory trouble. PayPal, on the other hand, shines.

Interest rates rule the day.

In 2023, every major fintech company is a quasi-bank that derives a sizable portion of its revenue from net interest income. coinbase, robinhood, and many others are reaping excess profits from rising interest rates.

Adding stablecoin to their offerings is enticing because it is similar in design to a zero-coupon bond. Stablecoin issuers issue tokens but get to keep all the proceeds they receive from short-term treasury bonds. In a perfectly rational world, no one would own stablecoins or cash without a constrained competitive market. What's the benefit of owning an asset that doesn't generate interest income? In short ...... none.

But we live in a world of huge regulatory hurdles, sanctions, foreign exchange exchange restrictions, domestic capital controls and securities laws. For some, there are more than enough ways to get dollars. In the cryptocurrency space, the need for leverage has far outweighed the need for imported Treasury yields until recently. With short-term yields above 5% and no signs of cooling, a new paradigm has emerged.

Jen-Hsun Huang noted that in order to meet the growing demand for generative AI, data centers need accelerated computing platforms that are tailored to specific needs. The new GH200 chip platform offers superior memory technology and bandwidth, the ability to boost lossless connected GPU aggregation performance, and a server design that can be easily deployed throughout the data center.

It is worth mentioning that the incoming wave of big models is spawning a variety of AI-native applications, driving a surge in arithmetic demand, and the data center market specifically designed to cope with data-intensive AI applications is rapidly emerging.

Data centers usher in new changes

As established cloud computing providers race to retrofit data centers with advanced chips and make other upgrades to meet the demands of artificial intelligence software, some upstart builders see an opportunity to develop new facilities from the ground up, analysts note, according to the Wall Street Journal.

A data center resembles a large warehouse equipped with multiple racks of servers, networking, and storage for storing and processing data. AI data centers have more servers using high-performance chips than traditional data centers, so AI data center servers can consume an average of 50 kilowatts or more per rack, compared to about 7 kilowatts per rack in traditional data centers.

This means that AI data centers will need to build additional infrastructure capable of delivering higher power, and because the additional power usage generates more heat, AI data centers will also need other cooling methods, such as liquid cooling systems, to prevent equipment from overheating.

Manju Naglapur, senior vice president at services and consulting firm Unisys, noted:

Purpose-built AI data centers house servers that utilize AI chips, such as NVIDIA's GPUs, and can run multiple computations simultaneously as AI applications sift through huge data stores. These data centers are also equipped with fiber optic networks and more efficient storage to support large-scale AI models. AI data centers are highly specialized buildings that require a significant investment of money and time. Spending in the global AI infrastructure market is expected to reach $422.55 billion by 2029, with a compound annual growth rate of 44 percent over the next six years, according to research firm Data Bridge Market Research.

DataBank CEO Raul Martynek said the pace of AI deployment is likely to lead to a shortage of data center capacity in the next 12 to 24 months.

AI arithmetic upstart secures $2.3 billion in funding

Currently, various giants are betting on AI data centers, and "real estate benchmark" Blackstone has sold its house to invest in AI data centers.Meta has also said that it will build a new AI data center.

As mentioned in the previous article, CoreWeave, an AI computing power newcomer, took NVIDIA H100 collateralized loan and obtained debt financing of $2.3 billion (about 16.5 billion yuan).

CoreWeave said the funds will be used to accelerate the construction of AI data centers, which is another funding round for the company after it received $221 million in April and $200 million in May this year.Founded six years ago, CoreWeave has seven AI data centers online and is expected to double by the end of this year.

CoreWeave is working with NVIDIA and Inflection AI to build a mega AI server cluster with the goal of running 22,000 NVIDIA H100s. if built, it would be the largest AI server cluster in the world.

It's worth noting that according to CoreWeave's official website, their services are 80% cheaper than traditional cloud computing vendors. NVIDIA's latest HGX H100 servers, which contain eight H100s with 80G of video memory and 1T of RAM kind of thing, start at as little as $2.23 (16 yuan) per hour.

And compared to its predecessor, the new GH200 Grace Hopper platform's dual-chip configuration increases memory capacity by 3.5x and bandwidth by three times, with 144 Arm Neoverse high-performance cores, 8 petaflops of AI performance, and 282GB of the latest HBM3e memory technology in a single server.

No wonder in this era of LLM explosion, Jen-Hsun Huang still boldly said "the more you buy, the more you save"!

In an interview on Monday, July 3, Bridgewater-connected Chief Investment Officer Greg Jensen systematically talked about Bridgewater's views on AI technology, sharing his thoughts on how Bridgewater invests in AI, how it uses AI investments, and its outlook on AI technology.

How Bridgewater invests in AI

Jensen states:

In restructuring Bridgewater, we also did something that we hadn't done before, which was to get some people to research and invest in things that might not be immediately profitable, but that were our long-term projects. So we set up this AI program with a team of 17 people, led by me. I'm still actively involved in the core of the Bridgewater fund, but the other 16 people are 100% committed to reinventing the Bridgewater fund by way of machine learning. We're going to set up a fund run exclusively by machine learning technology, which is what we're doing now in the lab and pushing the limits of artificial intelligence, machine learning capabilities. Now, trying to set up such a fund is still very problematic. If we use large language models, they have two types of problems. The first is that these models are trained more in the structure of the language, so they usually feed back something that looks like a structurally, grammatically correct answer, but not always an accurate answer. That's a problem. Second, it hallucinates, it makes things up because it's more concerned with the structure of the word or concept that comes up next than whether the concept that comes up next is accurate. Therefore, Jensen believes that AI can help people conceptualize and theorize what they observe, but there is still a long way to go to really use AI to select stocks. Thus, Bridgewater's real focus is on:

But there are other ways to combine that with statistical models and other types of AI. That's what we're really focused on, namely combining large linguistic models that are less accurate with statistical models that are good at accurately describing the past but are terrible at predicting the future. Combining these together, we are starting to build an ecosystem that I believe will enable what the Bridgewater analysts are doing. If this ecosystem is built, we will have the equivalent of millions of investment partners at the same time who are in the upper-middle range. If we have the ability to control AI illusions and errors through statistics, we can get a lot of work done quickly. That's exactly what we're doing in the lab and proving that the process works.

How does Bridgewater invest through AI?

If it were possible to build an ecosystem that encompassed AI and other technologies, how would Bridgewater use this system to make investments?

Jensen believes that statistical AI and large-scale linguistic models can complement each other and play a "right hand" role in Bridgewater's investments:

Statistical AI can take theories, go back and see if they were correct at least in the past, and what their flaws were, and refine them and provide advice on how to do things differently, and then we can talk to them. One advantage that large-scale language models have is to take a complex statistical model and discuss what it is doing. There are ways to train language models to do this. The way we model this situation is that the language model can come up with potential theories. It's not the most creative thing in the world, but it's a theory at scale, that's for sure. Again, large-scale language models are very good, but we have to somehow tweak the language model and we can then use statistics to control it. We can then use the language model again to get the results in the statistics engine and discuss them with humans or other AIs and report what was found, the content, and the type of theory. If the conclusions reached are contrary to what people perceive, then more testing is done. That's the loop I'm very excited about, and as I said, so far statistical AI has been limited because it focuses on market data. The benefit for linguistic models is that it can better understand things that statistical models don't. For example, statistical models of the market don't have the concept of greed, but large-scale linguistic models can understand the concept of greed almost as well - these models have read all the articles about greed and fear and so on. So combining the two now produces a human-like mode of thinking.

What does AI mean for human employees?

As time goes on, computers can do more and more. according to Jensen:

I would say that today, humans are used to only fulfilling roles related to intuition and creativity, and we use computers to memorize and run those rules with constant accuracy. That's only halfway through the transition, and now it's time to take another leap. There is no doubt that AI will change the role played by investment assistants. To be precise, we will still need people to work around these things for the foreseeable future, and it will still take us a while to build the ecosystem of these machine learning agents and so on. Leveraging AI is going to be part of the future of work, and I think it's hard not to leverage these technologies in any knowledge industry. In terms of writing computer programs, we're seeing a huge breakthrough in coding. Now, with AI, people just need to know what they want to code, rather than needing to know how to code, which is a huge breakthrough. So a bunch of people who are not well trained or competent in C++, Python, or whatever can suddenly get what they want much faster. So all of a sudden the skill sets that are needed in the workplace are changing and they're changing in ways that are surprising to many people because it's actually a lot of knowledge work like content creation and so on that people once thought was still in the distant future to be replaced by machines, but is actually close at hand. So the bottom line is that there are so many changes that it's essential to have flexibility in the workplace and to be able to take advantage of any tools that are available.

Can you use AI to manage investments directly?

With a variety of AI investment management tools now appearing in the market, the concern is that with the great development of AI, is it possible that in the future humans will only need to leave their investments to AI?

Jensen argues that:

I think it's both going to lead to accidents and it's very exciting to me. Obviously, I'm excited about the power of AI, and I think there are ways to make good use of it. But at the same time, AI can create a lot of mistakes. Some foundations use GPT to pick stocks, but these fund managers don't really have a deep understanding of AI and the weaknesses that can exist. In one example, in the real estate market, Zillow, a real estate brokerage platform, used AI technology to predict home prices, evaluate them, and enter the market to start buying houses that the AI thought were undervalued. However, Zillow has several problems. One is that while they have a large amount of housing data, that data is happening over a relatively short period of time. So even though they have what appears to be a large number of data points, there is still a macro cycle that affects the assessments that they make. Secondly, they underestimated the disconnect between theory and practice when it was in fact an adversarial market. So, this is obviously a huge problem for Zillow, who had a big impact on the real estate market and then suffered a huge failure. Going back to the stock market, very short-term trading is arguably better suited to machine learning because there is a lot of data through which AI can learn faster. But on the other hand, in the longer term, AI may not be able to play a role. The data is usually like a person's heart rate data over a lifetime. You might think, wow, my heartbeat has been going on for 49 years, and that seems like a lot of data, but when you have a heart attack, that data is completely irrelevant. So even with a lot of data, it can be misleading, and those issues will lead to huge problems with these technologies. So one has to understand these tools, what they are good at and what they are not, and put them together in a way that plays to the strengths of each type of tool and circumvents the weaknesses. There is still a lot of work to be done on large language models, and we can certainly train them through reinforcement learning to make sure they don't make known mistakes.

Are markets still dominated by optimism?

Jensen believes that the market is still dominated by optimism. He said:

The Fed seems to be a little more realistic than the market in terms of the actions it will take. When you look at the market's reaction, you see that it's very optimistic. But we have to note that historically, the market is often prone to over-optimism.

On July 3, Titanium Media App learned that Li Kaifu's latest AI large model startup and official website was officially unveiled, the company is called "Zero One Everything" (01.AI), headquartered in Beijing.

It is reported that the zero-one everything is positioned to build AI 2.0 platform and application, the technical route for self-research large model, within three months has achieved 10 billion parameters scale model internal testing, is now expanding to 30 billion - 70 billion parameters scale. The team members who have joined Zero2 Everything include former Alibaba vice president, former Baidu vice president, former chief algorithm leader of Drip/Baidu, former Google China executives, etc. "Zero One Everything is recruiting AI talent from around the world.

The information in the sky-eye search shows that the main business entity of Zero One Everything is "Beijing Zero One Everything Technology Co. The company's shares are 99% owned by Ma Jie, the former vice president of Baidu and the person in charge of the product "Xiyang", and 1% owned by Innovation Works (Beijing) Enterprise Management Co.

Titanium Media App learned that although Ma Jie holds 99% of the shares and is the executive director of the company, but Ma Jie is not the CEO of Zero One Everything, but only an executive, and the Innovation Works team is acting as the co-founder of the company, and Li Kaifu will not be All in this company, and he will balance the company and the fund project afterwards.

At present, zero one million things did not announce more core team information.

Earlier during the 2023 Zhongguancun Forum, Li Kaifu told Titanium Media App and others that the generative AI fever represented by GPT-4 and other big models is spreading around the world, meaning that the AI 2.0 era has arrived, which will bring opportunities ten times greater than the mobile Internet era, penetrating all walks of life to greatly boost productivity. "We predict that the new AI 2.0 platform will effectively help boost human productivity and create tremendous economic and social value."

On March 19 this year, Kai-Fu Lee announced his entry into the AI big model field, personally organizing the new Project AI 2.0 company, which he said is a global company dedicated to building a new AI 2.0 platform with AI-first productivity applications. "It's an innovative company driven by a technology vision, with a superior Chinese engineering heritage, and a global call for world-class talent to join."

Three months later, the seventh company incubated by Lee Kai-fu's Innovation Works, Zero One Everything, went live.

Li Kaifu said, zero-one is 01, which represents the digital world, from zero to one, and even everything in the universe, implying the ambition of "zero-one intelligence, everything empowerment". Innovation Works said that this is Li Kaifu "rolled up his sleeves" personally led the team, "the tip of the tower incubated" an AI 2.0 company, after three months of preparation period.

"The technical threshold of large models is very high, and it is very necessary to integrate an efficient team with both scientific research capabilities and development strength, to develop and implement the company's technology and product strategy, and to implement it consistently from top to bottom." According to Kai-Fu Lee, Zero One Everything will focus on building a platform and productivity application ecology for the AI 2.0 era, and the vision of Innovation Works is to build Zero One Everything into a world-class company that shows our R&D and engineering capabilities to the world.

Specifically, in order to build AI 2.0 big language models, Zero2 One Everything chose to "self-research models" and build an experimental platform - building thousands of GPU clusters for training, tuning and inference to increase the number of effective data parameters, as well as the density of high-quality data used. We also built an experimental platform - a cluster of thousands of GPUs (graphics processors) for training, tuning and inference, to increase the number of valid data parameters and the density of high-quality data used to build continuous, controlled and scientific training capabilities.

In order to improve the modeling capability, Zero2 Everything adopts a strategy of dismantling the technical route and attacking it step by step, including 7 modules:

Pre-training framework (Pre-train) technology that defines the ceiling of model capability; Post-train (post-training) technology that refines the model realization capability; AI Infra technology that supports efficient training of models; Multimodal modeling techniques for more diverse and multi-media; Platform middleware and toolchain technologies; Inference deployment technologies; Service deployment technology. At present, the technical experts of the seven directions of Zero2IPO come from the background of many Chinese and foreign technology companies, with a high proportion of PhD background, and have achieved the internal testing of models with 10 billion parameters in three months, and are expanding to 30 to 70 billion parameters. At the same time, Zero2IPO also plans to launch an open source version of the model in the future to attract outstanding talents to enhance the model and promote the global AI 2.0 developer ecology.

"It's not a Chinese or American problem. I think all companies will tend to think about their core business and how they can integrate with AI." For arithmetic, there are other chips available in China as AI arithmetic alternatives, Kai-Fu Lee told Bloomberg in March this year. "In any case, it's not a desperate situation."

Li Kaifu has said that the main focus of Innovation Works in the AI 2.0 era is on three major directions: intelligent applications, platforms and infrastructure. With a cutting-edge technology layout of "tower incubation + angel investment + VC fund", Innovation Works will actively support hard-tech entrepreneurs in AI 2.0 and help China continue to lead the world in the next golden decade of AI.

Vision Pro is very cool, and from the overwhelming reviews, we also see Apple's strong strength in hardware and UI design, and aim at the possibility of a future generation of computing platforms, but while excited, in this whole conference, Silicon Star people are also paying close attention to another important topic - in the current rolling wave of AI What is Apple doing and what does it intend to do with the current wave of AI?

And if we're just talking about AI, this event was a big disappointment mixed with a little hope. Some expectations are that although Apple did not mention the word AI this time, important AI concepts such as machine learning and Transformer models still appear from time to time, and a number of its products have been updated with some AI features, in addition to which the launch of the MR headset may also open up a new imagination of Apple's AI applications.

However, the whole conference did not launch a generative AI-based self-developed products, and even next door Microsoft has begun to enable AI assistants at the system-wide level, Apple is still playing the old game of speech-to-text, keyboard auto-correct fill - which is inevitably a little disappointing. Don't forget, it was Apple's Siri that started the era of voice assistants. But in today's ChatGPT sweeping the world, Apple's Siri just a less "Hey"?

At least from this year's WWDC, Apple has made an effort on AI, but not much.

WWDC, what AI updates did Apple unveil this time?

Unlike Google IO and Microsoft Build, which just ended in May and almost all revolved around AI, the AI at Apple's developer conference was more like an egg-like presence, buried in hardware and software updates. Although it brought some product optimization, it did not make people's presence shine.

First, at the hardware level, the newly released M2 Ultra chip enables up to 192GB of unified memory by combining two M2 Max chips together, 50 percent more than the M1 Ultra, thus enabling smooth support for large model computing and accomplishing tasks that other chips cannot. For example, Apple said the M2 Ultra can be compared to other PC devices in support of Transformer large models run on the advantage of obvious.

In addition to the chip, AirPods have also quietly introduced some new AI-powered features. For example, AirPods now have "adaptive audio" capabilities that automatically adjust the volume of media playback based on your environment or remember your preferences at different times, allowing you to focus on content or communicate with people, backed by machine learning technology.

In addition, along with the OS upgrade, some useful little AI features are becoming available to Apple users. But to be honest, there are no highlights and it all gives a feeling of déjà vu.

First of all iOS 17 this time added a new native application called Journal, similar to the previous Memory upgrade. Based on machine learning technology, the iPhone is now able to intelligently mark and record significant events or interesting moments in a user's life, and automatically add details to any entry for photos, music, recordings, and more to make it easier to go back.

In addition, Live Voicemail (live voicemail) was launched, and both voice messages and others' iMessage voicemails can now support real-time voice-to-text conversion. But this feature is all too familiar to the majority of Chinese users - after all, WeChat users started using this feature years ago, and not long ago OpenAI also open-sourced its audio-to-text model Whisper, which can accurately identify 98 languages. This time Apple did not specify what model was used to develop the feature, only that it was based on a powerful neural network engine.

The iPhone's own keyboard now has the ability to automatically predict word recommendations and corrections. Apple says that now every time we press a key, the iPhone will run the Transformer language model under the powerful computing power of Apple's chip.

While people are still amazed at the ability of AI paintings to be fake, AI voice imitation technology has been applied by scammers. In Canada, at least eight victims were cheated out of about $1 million in just three days, with scammers imitating the voices of their loved ones to "save the day", even if their own mothers couldn't hear the mistake.

With AI, the "stingy man" can become a seductress in seconds, strangers can easily impersonate your friends and relatives, and you will lose if you play money.

From blockchain to AI, every time a new technology comes out, fraudsters are the first to charge the frontier of the times. In the face of fake scams, ordinary people are increasingly difficult to prevent, those who open Pandora's Box of technology giants, but also can not just be busy running blindfolded, have to use magic to defeat the magic.

AI imitation voice scam emerged 8 people in Canada were cheated 1 million

"Mom! Help me!" The cry of her daughter came from the other end of the phone, and the mother panicked at once. Before you can get over it, you can hear an unfamiliar man on the phone yelling, "Tilt your head back and lie down. The next strange man spoke, "Your daughter is here with me, you dare to call the police or notify anyone, I will stun her and take her to Mexico."

At the end of the phone, the anxious mother could faintly hear her daughter's cry for help, "Mom, please help me, help me." At this point, the "kidnappers" talked about the conditions, the need for $1 million to release, not so much money in response, the "kidnappers" changed his words, "take $50,000 for your daughter. "

The above scenario is a real scene that happened recently in Arizona, USA. According to NBC, this is an AI imitation voice scam. The girl's voice on the other end of the phone was, in fact, an AI clone. Fortunately, the mother calmed down and contacted her husband to confirm that her daughter was okay before she was saved from being scammed.

When recalling the details of the call afterwards, the mother felt a wave of fear, although it was clear that this was a telecom fraud, but she still did not have the slightest doubt about her daughter's voice.

At a time when artificial intelligence is speeding up technological change, unscrupulous elements are using AI as magic to commit fraud with the most advanced technological means. How advanced is this technology? According to Kambam Patti, a professor in AI at Arizona State University, it is possible to quickly replicate the exact same voice by simply obtaining 3 seconds of someone's original voice.

Many people have already been cheated. on April 2, CCTV released a video report that Canadian criminals used AI to synthesize the voices of their loved ones to commit fraud, and at least eight people fell for the scam in three days, involving about 1 million yuan. Most of these victims were elderly people, and their feedback was that "the voices on the phone were simply identical to their loved ones."

Even the "voice print" system, which was previously considered secure, could be easily cracked. In March, Guardian Australia reporter Nick Everside said he used artificial intelligence to make his own voice and successfully accessed his Centrelink self-service account. And just two years ago, the Australian Service claimed in a report that "voice prints are as secure as fingerprints".

The speed of AI evolution is too fast. Back in 2019, the Wall Street Journal reported on a scammer who used AI voice synthesis to impersonate his boss and get the managing director of a branch to pay him 220,000 euros. But at the time, the scammer also had to prepare a corpus of the impersonated person to be trained and polished. Now, AIGC makes it possible to clone voices with almost no threshold, even dialect and tone of voice imitation are not a problem.

At present, there is no large-scale case of AI voice imitation fraud in China for the time being, however, the use of AI to generate fake pictures to cheat money has happened frequently.

In February this year, a "yacht maid party will be held in Suzhou Jinji Lake" news circulating in the network, publicity documents with a number of maid dress up female photos, the publisher explicit price, to participate in the Party each person to pay 3000 yuan. The blatantly organized "erasure" party, soon attracted the net police, according to the follow-up net police disclosure, those maid photos are basically AI generated.

There are many similar incidents, and netizens have gradually developed a sense of precaution that "there is not necessarily a picture of the truth". Nowadays, even if there is sound, there is video is not credible.

Anti-forgery technology can not keep up with AI "high imitation" anti-fraud to be upgraded

When ChatGPT was asked "What types of AI fraud are there?", it gave several answers, including using AI to generate fake websites and social media accounts, impersonating legitimate organizations and individuals to obtain information and money; using AI to generate simulated voices to commit wire fraud; deep learning of victims' personal information to generate personalized fraud schemes And so on.

But ChatGPT still underestimates the tactics of scammers. For fraudsters who set up elaborate schemes, AI serves as the underlying layer and even gives rise to an endless number of fraudulent methods.

In this regard, "piggy banks" that incorporate AI technology are not less harmful. On various dating and matrimonial sites, there are many "piggy bank" scams, in which crooks fake their own identities and personas to get close to their victims, and then gain their trust and obtain money.

In the past, scammers would steal photos on the Internet to disguise themselves, but as long as there are people who are interested in searching the picture comparison may also be detected, but if the scammers use AI to draw, this can be a "unique". Not only that, through AI drawing, the scammers can even easily make corresponding gestures according to the victim's request, making it easier to gain trust.

On social media, it is also easier for "stingy men" to pretend to be beautiful women to cheat money. Some AIGC applications can generate seductive, large-scale photos and videos to entice users to chat naked or make them pay for more intimate photos. Put in the past, these visual products still need to spend efforts to produce, now these things are directly handed to AI, once someone steps into the trap, they will be taken by the scammers.

What's even scarier is that since many AI tools have been open-sourced, the cost and threshold for scammers to commit fraud is getting lower and lower, and tutorials are available directly on the web. The publishers' intention is for people to learn, but they can't stop unscrupulous people from getting their perverse ideas.

With the technological evolution of AI, what people hear and see will become more and more untrustworthy. To avoid being cheated, not only do we need to understand fraudulent methods and raise awareness of prevention, we also need advanced technology to defeat magic with magic.

In the face of the harm brought by depth forgery, Google began to consciously defend as early as 2019, first recruited a group of actors to record videos, and then use the online depth forgery technology to change faces and other operations, and to analyze these data comprehensively, so that developers can study the algorithm and logic of depth forgery technology, in "know yourself and know your enemy On the basis of "know your enemy", more accurate identification and prevention of deep forgery content.

Last year, Microsoft launched a tool called "Video Aunthenticator", which can analyze rendering boundaries and grayscale levels on a frame-by-frame basis in real time to generate a reliable index to help users identify the authenticity of content.

Not long ago, OpenAI also released an AI-generated content recognizer, but blog data shows that the detection success rate is only 26%, and the detection effect is much worse for languages other than English. Compared with the power of ChatGPT, the AI company's research on "counterfeiting" is not deep enough.

When AI technology presents two sides of the coin, the battle between good and evil can never be stopped. It is thought-provoking that in 2018, the State University of New York had developed an "anti-facial" AI forensic tool, which identifies forged faces by predicting whether the eyes are blinking or not, with an accuracy rate claimed to be 99%, but the tool soon failed because the deep forgery technology evolved.

The world needs them to master the chains that hold the devil in place as the technology giants open Pandora's Box.

Imagine a world where you can fully immerse yourself in an interactive and captivating game. You are no longer a passive observer but rather an active participant in the game's world, where every action you take has consequences. Our game will take you on a journey through a fantasy world filled with breathtaking landscapes, mysterious creatures, and epic battles.

The game will be designed to give players a sense of freedom and control over their gaming experience. The world will be open and vast, allowing players to explore and discover new areas and challenges at their own pace. Whether you want to embark on a quest or engage in fierce battles, our game will cater to your interests and offer a multitude of options.

The game will feature state-of-the-art 3D graphics that will make you feel like you are living in a real-world rather than just playing a game. The creatures and characters you encounter will be intricately designed with stunning detail, and the world's landscapes will be breathtakingly beautiful.

To ensure that players are always engaged and challenged, the game will feature an extensive leveling system that rewards players for their progress. Whether you are a casual player or a hardcore gamer, you will find the game's challenges to be just the right level of difficulty.

One of the most exciting features of our game is the integration of NFTs. As a player, you will have the opportunity to mint your own NFT, which will give you access to exclusive drops and rewards. This feature will give players a sense of ownership and investment in the game, making it a truly unique and personalized gaming experience.

Players who subscribe to the game's account will receive a free NFT and access to future drops, ensuring that they have a vested interest in the game's development and success. As the game grows and evolves, players will have access to new content, weapons, and features that will keep them engaged and invested in the game.

<subscriber-edition://0x7cF3FeA43E0E1546aF670B3062F8015aEDeee544?tokenId=0

In addition to the game's core features, we will also be offering a range of exciting mini-games and challenges that will allow players to earn rewards and progress through the game. These mini-games will be designed to be quick and accessible, making them perfect for players who want to take a break from the main game and still make progress.

To ensure that the game is always up to date and relevant, we will be constantly releasing new content, challenges, and features. This will keep players engaged and invested in the game's ongoing success, making it a game that players will want to keep coming back to for years to come.

In conclusion, our upcoming 3D game is set to become the game that sets the standard for all future games to follow. It offers an immersive and captivating gaming experience that will leave players wanting more. With stunning graphics, intricate design, and unique features such as NFTs, this game is sure to be a hit with gamers around the world. So why wait? Join us on this thrilling journey and discover a world unlike any other.

Crypto startups are looking to expand into new banking after the sudden collapse of three prominent crypto banking partners, Silvergate Bank, Silicon Valley Bank (SVB) and Signature Bank, which includes alternative banking services such as Mercury, Brex and Customers Bank.

Venture capitalists at a leading cryptocurrency investment firm said that while depositors had little to lose, the closures of Signature and Silvergate Bank - which also had to cease operations last week due to a bank run - have had a serious impact on the crypto sector. It's a euphemism for the fact that there are fewer and fewer banking options in the U.S. for projects in the space, which isn't really good news," they said.

Operational bottleneck 2.0

The investor urged the portfolio company to add at least one more bank, but finding one is easier said than done.

Since the end of last year, U.S. regulators have stepped up their regulatory actions against crypto companies, especially those associated with banking services, whether it's a stablecoin issuer like Paxos or a digital asset bank like Custodia. Venture capitalist Nic Carter has dubbed this "Operational Bottleneck 2.0," referring to the crackdown on cryptocurrencies through the banking sector.

Carter said in a tweet, "A month ago it was a bottleneck - banks discouraged exposure to cryptocurrencies, now it's almost certainly ...... decimated, and every bank that supports cryptocurrencies has closed."

George Bousis, co-founder and venture capitalist at early-stage venture capital firm Protagonist, remains cautious in recommending banking partners, as the landscape changes daily.

I wish I had a better answer, like having a crystal ball to know what's going to happen at this point, but I don't, and that's one of my concerns and something that keeps me up at night," he told The Block.

What are the options for crypto startups?

While publicly cryptocurrency-friendly banks in the U.S. are fading from view, some web3 companies are still trying to find banking partners.

Options include Mercury, Brex and even Wall Street giant JPMorgan Chase, said the venture capitalist, who asked not to be named. JPMorgan declined to comment, and Mercury and Brex did not immediately respond to requests for comment.

In the medium term, we prefer the big banks [such as] Chase, but at this point, it does suggest there is any plan B or C, so a platform like Brex or Mercury that can work together quickly is a good option," the anonymous investor said.

Den, a multi-signature startup, has been using Mercury as a banking partner in conjunction with on-chain assets.

Speaking about the Mercury partnership process, Den co-founder Jonah Erlich said Den is taking advice from venture capitalists looking to increase the percentage of funds it holds on the chain following the weekend's breakout. It's been a very smooth process for us and we're staying with Mercury, but also exploring opening accounts on similar fintech platforms using clearing accounts and GSIBs [global systemically important banks]," he said.

Mercury is a fintech startup that works with two FDIC-insured banks, Evolve Bank and Trust and Choice Financial Group. Its checking and savings deposits can now be insured by the FDIC for up to $3 million, compared to the standard $250,000 threshold. This is thanks to the use of sweep networks, which allow customer deposits to be distributed across banks.

We had twice as many new customer additions today as we do on a normal business day," said Immad Akhund, CEO and founder of Mercury, over the weekend.

Another option in the U.S. is Customers Bank, but many of these bankers do not have the same access to deposits that Signature and Silvergate offer, said Michael Safai, co-founder and partner at Dexterity Capital, a trading firm.

Regional and community banks

Another challenge is that none of the large U.S. banks or even regional banks support holding tokens or crypto assets.

Mercury itself is privately held, but its partner, Choice Financial Group, trades it publicly. Despite U.S. President Joe Biden's assurances on Monday, the stocks of many regional and large public banks have also fallen severely in the wake of recent events.

Neil Zhang, founder of prediction market startup Frontrunner, has had trouble securing bank partnerships due to the double whammy of running a web3 company that focuses on online betting. His startup now relies on two regional banks, MVB and Western Alliance, neither of which immediately responded to a request for comment on its product.

It took us a long time to find two banks that would take us on, and while they're not the newest or most technically advanced, we've been happy with their service since switching from Brex earlier this year," Zhang said.

He added that the company does not intend to abandon either bank, but has been looking to diversify further, despite the fact that both banks have taken significant hits in the open market.

Assets on the Chain

Many crypto startups store at least some assets on the chain, Zhang said, adding that Frontrunner primarily holds legal funds, but crypto assets are stored in the multi-signed Gnosis Safe wallet. The startup hasn't received any guidance from investors who suggested it change that approach, he added.

Thibaut Sahaghian, CEO of Multis, a Web3 company card startup, said Multis has also not received any advice from investors and is simply allocating funds to multiple banks. He added that it is still attracting many web3 customers because it is a viable alternative to Mercury's offering of checking accounts and corporate cards.

As of now, Multis is in a good financial position," Sahaghian said. We are fortunate to have limited exposure to SVB and we still have sufficient funds in multiple accounts (including ours) in USD, EUR, USDC, DAI, ETH and BTC," said Sahaghian.

The startup currently uses Synapse, which works with Lineage Bank, for banking purposes and does not plan to change providers. Synapse and Lineage Bank did not immediately respond to a request for comment on their products.

We are not immune to moral hazard and therefore need to diversify into various assets to keep the business running, and USDC and stablecoins are only part of the answer, as redemptions have been affected by the current crisis, by relying on centralized issuance of fiat currencies," he said. But we are moving in the right direction."

Eyes turn outside the U.S.

Startups not based in the U.S. have the advantage of taking advantage of players like Sygnum and Seba Bank - Swiss banks that are also allowed to hold crypto assets.

Sygnum Chief Client Officer Martin Burgherr told The Block that the bank received a large number of inquiries following the impact of Silvergate, but they were unable to fully fill the void due to the company's strategic decision not to serve U.S. clients.

Crypto companies looking for new banks may have to look overseas for the time being," said Safai, co-founder of Dexterity Capital. There are several banks in the Bahamas, and there are some cryptocurrency-friendly and trustworthy options in Europe." Translated with www.DeepL.com/Translator (free version)

With this in mind, what is a DAO (Decentralized Autonomous Organization)? In my experience so far, DAOs are online discussion groups that have, at best, a shared pool of funds and a governance system to help determine how to use those pools. But the real promise of DAOs is that they can be programmable organizations in the same way that cryptocurrencies are programmable currencies.

The concept of programmable organizations brings with it the ability to create composable, reusable organizational structures and capabilities.

While DAOs are ostensibly made up of people (or other DAOs, bots, or ......), at their core, DAOs are different from any other population in that they are managed through code. In programmer-speak, we interact with DAOs through APIs or application programming interfaces. In decentralized blockchain-based systems, these APIs take the form of functions that can be called publicly on smart contracts. While the DAO is programmed by humans (and sometimes other programs), the real functionality of the DAO is its smart contracts. Everything else is, as the Orca protocol team likes to call it, SocialWare (a system based on human trust run manually by humans).

DAOs are essentially programmable components that perform functions through APIs.

What are these capabilities? What is an API?

In terms of programmable systems, right now, most DAOs consist of two things.

Membership lists Shared wallets / vaults Discussions happen in Discord, meetups happen at conferences, groups start writing code, designing things, tweeting, etc., but the essence of most DAOs - to distinguish them as DAOs from online clubs - boils down to these Two functional areas. Today's DAOs implement code to manage memberships and spend money.

So what does it mean that the components that can manage members and wallets are composable?

Publicly accessible functionality for membership management (joining, leaving, etc.) and multi-party wallets (voting, executing transactions, transferring funds). Standardized interfaces for these functions (names of records, inputs and expected outputs) so that multiple implementations can be interchanged. A standardized set of invocation conventions, so that each set of functions or interfaces does not require a completely different interaction. This has the advantage of allowing different types of modules to be substituted for each other or managed in the same way. It also means that calls can be linked together in a way that the interface creator did not plan. A historical example of a calling convention like this is that tools for UNIX-based systems almost always read text from their input and output text, thus allowing them to be linked together with pipes. The Zodiac team has done some excellent work in creating proposed standards for composable DAO behaviors, similar to the conceptual approach of aspect-oriented programming.

While we are now discussing composability in the context of behavior, composability also applies to data. For example, the Orca protocol implements a DAO consisting of many DAOs (called Pods). each DAO can have zero or more DAOs as parents in a graph of theoretically infinite size and complexity. At the top level, a DAO consists of all its members recursively, whether they are humans, DAOs/Pods, or other identities associated with bots, etc.

Now that we have a simplified view of what most DAOs are and what composability means for them, you may think it doesn't sound exciting that the world of human collaboration is far richer than simply the ability to form teams, raise and spend money. The truth is, these People Lego's we've built so far aren't very useful. What else do they do but exist? Not much, and while the humans in DAOs may be busy and very productive, creating art, music, companies, online materials, or trying to save the planet, DAOs themselves are superficial and almost featureless.

The next step in DAO composability is to extend the functionality of the DAO itself, and to extend the surface area of the DAO API accordingly.

My approach is to start by implementing a concrete instance, iterate over that instance, and then move to another instance. After some concrete implementation, the pattern emerges. Let's explore one possible example: customer service.

Imagine that a new DAO, SupportDAO, is created to service customer support tickets for a Web3 company, which might be a for-profit organization whose members are customer support experts and have built a discipline around handling online customer support at scale. metrics for customer support, almost all of which are easy to measure (see examples below).

SupportDAO has members and a shared vault, so it can implement standard interfaces for these high-level functions in a composable generic DAO ecosystem, but what else can SupportDAO do? At the core of its API is the ability to provide customer support tickets from a ticketing system to it to resolve those tickets and then make payments based on the successful resolution of those tickets.

How does a DAO (or company) pay another DAO for processing customer support tickets without trust? Payments are made based on an agreed set of KPIs. For example, the payment might be for the following functions.

Number of tickets processed Time to resolve the issue Time to first contact Percentage of resolved work orders versus unresolved closed work orders Average age of queued work orders Customer NPS etc. The fact that these are measurable and can be integrated either on-chain (probably not) or via a prophecy machine means that SupportDAO can run as an autonomous, trustless component rather than being manually orchestrated by humans for payment, which is an admittedly naive and superficial view of the design, but I find it exciting.

A DAO-enabled API like this would be composable, and any organization building an application or service could evaluate plugging it in and having new customer support capabilities, rather than building its own customer support organization, and more importantly, if the API is standardized (or even de facto), multiple supporting DAOs could compete in real time to provide better, faster service to each organization while (or even anonymously) sitting behind the API.

Customer support is just one of the areas that can be achieved through smart contracts and standardized APIs. As we explore the future of decentralized organizations and collaboration models that look more like federations of small teams than monoliths, more interfaces and automated collaboration mechanisms will emerge.

With the implementation of digital RMB application scenarios and the exciting offers of Digital RMB Consumption Week, payment with digital RMB has quietly become a new fashion in Xiamen's consumption market during this year's May Day holiday.

During the "May Day" Digital RMB Spending Week, major banks and market players in Xiamen have come up with preferential packages to accelerate the popularization and promotion of digital RMB, with ICBC, Agricultural Bank of China, Bank of China, Construction Bank, Bank of Communications, Post and Reserve Bank, China Merchants Bank, Meituan, Jingdong, etc. all offering a number of digital RMB spending activities for Xiamen users. The financial empowerment has injected a lot of energy into the festive season.

After the opening of the "May Day" Digital RMB Consumption Week, many market players actively took the "Finance + Consumption" express and opened digital RMB payment services.

"It's convenient to use digital RMB for consumption, just move your fingers on your cell phone, it's convenient and safe, and there are unexpected exclusive consumption promotion offers." In front of the cashier of Wanxiangcheng, Mr. You quickly completed the digital RMB payment through the digital RMB wallet of Agricultural Bank of China, and at the same time, he enjoyed the special offer of Wanxiangcheng Agricultural Bank's "May Day" digital RMB consumption week by spending RMB1,000 and getting RMB100 off.

The Bank of Agriculture's Digital RMB "Coins with Gold" lucky draw; Bank of China's Digital RMB Meituan Gift Pack; CCB's Digital RMB Parking, Digital RMB Money Saving Card, and Digital RMB Food & Beverage Superstore Coupons; Bank of Communications' Million Digital RMB Gift Pack; Post and Reserve Bank's Digital Wallet New Customer Discount; and China Merchants Bank's Digital RMB eCard Koi for the whole year. The digital RMB e-cartoon carp is free for the whole year; Meituan digital RMB 40 gift pack; Jingdong digital RMB 110 gift pack ...... During the holiday period, Xiamen launched the "May Day" digital RMB consumption week, issuing digital RMB vouchers, red packets and lucky draws through multiple channels and scenarios. The company is also a member of the "Digital RMB" group.

The company's products and services are available in a wide variety of formats, including the Internet, the Internet, and the Internet. In Yuanchu Food, digital RMB activities were fully carried out, and "digital RMB" payment became a popular payment method in front of the cashier counter. The reporter saw that Yuanchu food together with CCB Life launched a big discount, CCB Life App newly registered users over 31 yuan can use 30 yuan digital RMB exclusive coupons, digital RMB users over 30 yuan can use 10 yuan (over 10 yuan can use 5 yuan) restaurant super coupons, and super spending over 15 yuan randomly 1-9 yuan and other activities, convenient experience superimposed on the exciting benefits, many people scrambled to Many people are eager to experience the digital RMB. According to the person in charge of Yuanchu Food, more than 10,000 people have used digital RMB to spend at Yuanchu Food since the campaign was launched.

According to industry insiders, the rich application scenarios of digital RMB will further boost the digital quality of Xiamen's business district and strengthen the foundation of Xiamen as an international consumption center city. According to the relevant officials of the Municipal Bureau of Finance and the Municipal Bureau of Commerce, Xiamen will continue to enrich the application scenarios of digital RMB in consumption, guide more entities to embed digital RMB payment methods, encourage multi-sectoral and cross-sectoral synergy to launch more digital RMB consumption promotion activities, create a number of digital RMB featured business districts, take the digital RMB pilot as an opportunity to further promote the quality of traditional consumption and upgrade, and stimulate consumption To further promote the quality and upgrade of traditional consumption and stimulate the potential of the consumption market by taking the opportunity of digital RMB pilot. *** Translated with www.DeepL.com/Translator (free version) ***