Core Viewpoint - The U.S. Department of Justice (DOJ) is investigating Google's technology transactions with Character.AI to determine potential antitrust violations, highlighting ongoing scrutiny of major tech companies like Google [1][2]. Group 1: Investigation and Background - The DOJ's investigation into Google is part of a broader focus on antitrust issues, with previous lawsuits against Google for monopolistic practices in search engines and advertising [1]. - Character.AI, founded in November 2021 by former Google AI team members, has gained significant attention in the AI sector, particularly after the launch of OpenAI's ChatGPT [2][3]. Group 2: Character.AI's Growth and Challenges - Character.AI's user engagement surged from 18 million visits in December 2022 to 500 million by March 2023, a 27-fold increase, following the rise of generative AI [3][4]. - Despite its popularity, Character.AI faces financial challenges, having raised $150 million in Series A funding but still requiring additional capital to sustain its operations [5]. Group 3: Google's Investment and Talent Acquisition - In August 2024, Google invested $2.7 billion in Character.AI, allowing it to use the company's language model technology while facilitating the movement of key personnel back to Google [5][6]. - The arrangement is viewed as a form of indirect acquisition, raising concerns about the potential decline in Character.AI's innovation and operational capacity due to the loss of its founding team [6][7]. Group 4: Implications of Talent Acquisition - The talent acquisition model used by Google may circumvent traditional antitrust scrutiny, as it does not involve outright acquisition but rather a strategic partnership [7][8]. - This approach reflects a trend among tech giants to secure talent and technology from startups without triggering regulatory challenges associated with mergers and acquisitions [9][10]. Group 5: Regulatory Considerations - The DOJ's investigation into Google's actions may not lead to direct legal action due to the complexities of current antitrust laws regarding talent acquisition [20]. - There is a call for regulatory updates to address the nuances of talent acquisition deals, ensuring they do not undermine competition or innovation in the tech industry [21][23].

Financial Performance - The company achieved strong revenue growth of 8.9% year-on-year, with EBITDA rising by 13.7% [5][11] - Adjusting for the deconsolidation of TNS Plus, revenue growth would have been 11.7% and EBITDA growth would have been 15.5% [6][12] - In local currency terms, revenue increased by 12.9%, outpacing inflation and nominal GDP growth [10][11] - Direct digital revenues grew by 50.2% year-on-year, now representing 14.3% of total revenues, up from 10.4% a year ago [6][18] Business Segment Performance - The Telecom and Infrastructure segment contributed $880 million, growing 4.2% year-on-year [13] - Direct digital revenues in Pakistan surged 49.5%, contributing 27.7% of total revenues in that market [21] - In Ukraine, total revenues grew 49.5% year-on-year, with direct digital revenues increasing by 141% [24] - Kazakhstan's telecom and infrastructure revenues grew 9.4% year-on-year, while EBITDA was impacted by the absence of regulatory tax benefits [26] Market Performance - Pakistan's revenue growth was 20.3%, with EBITDA growing at 13.2% [19] - Ukraine's operations achieved 20.2% revenue growth and 10.2% EBITDA uplift after adjusting for the previous year's cyberattack [20] - Uzbekistan reported revenue growth of 13.1% and EBITDA growth of 16.5% [30] Company Strategy and Industry Competition - The company is transitioning into a services-oriented business model, leveraging digital services and AI integration [6][7] - The asset-light strategy is a cornerstone of value creation, with ongoing initiatives to unlock infrastructure and tower value [7][8] - The strategic partnership with Engrave Corporation in Pakistan is expected to unlock $563 million in value [7] Management Commentary on Operating Environment and Future Outlook - Management expressed confidence in achieving full-year guidance, projecting 12% to 14% underlying local currency revenue growth [42] - Inflation rates in the company's markets are trending lower, which is seen as a potential headwind [42] - The company is optimistic about the future, expecting to continue delivering on its strategy and creating sustainable value [44] Other Important Information - The second phase of the share buyback program commenced, with $23 million of shares repurchased [8] - The company repaid $472 million of bonds in April 2025, with no additional maturities until 2027 [40] - The company is making progress on the proposed NASDAQ listing for KSTAR, expected to complete in Q3 [83] Q&A Session Summary Question: What percentage of revenue from multiplayer customers is due to their use of apps versus telecom revenues? - Management noted that multiplayer customers generate 3.7 times more ARPU compared to voice-only customers, with data consumption three times higher [51][52] Question: How does the company position itself against Netflix in Pakistan? - Management sees Netflix as a complementary service rather than a competitor, emphasizing the relevance of local content [53][54] Question: What is the strategy around data centers? - The company plans to partner with investors for next-generation data centers while maintaining an asset-light strategy [62] Question: What are the drivers behind ARPU growth in Ukraine? - ARPU growth is attributed to fair value pricing and the stability of the customer base, with tower assets separated from the operating company [72][73] Question: What is the guidance for CapEx in Ukraine? - CapEx in Ukraine has increased due to extraordinary investments in energy sustainability, but is expected to align with guidance moving forward [81][82] Question: Will the company roll out fintech products in other markets? - The company is focusing on regulatory issues to introduce financial services in Bangladesh and Ukraine, following success in Pakistan [91][92]

西风 衡宇 发自 凹非寺 量子位 | 公众号 QbitAI 怎么老是你？？？ （How old are you） 这是最近网友不断对着 Transformer八子之一的Noam Shazeer （为方便阅读 ，我们称 他为沙哥） 发出的灵魂疑问。 尤其是最近Meta FAIR研究员朱泽园分享了他们《Physics of Language Models》项目的系列新进展后，有网友发现，其中提到的3-token 因果卷积相关内容，沙哥等又早在三年前就有相关研究。 是的，" 又 "。 因为你只要梳理一遍他的工作履历，就不难发现，AI界大大小小的突破背后，总是能发现他的名字。 "不是搞个人崇拜，但为什么总是Noam Shazeer？" △ 网友称右下角沙哥图由GPT-4o生成 朱泽园也自己也站出来表示，沙哥成果超前： 我也觉得Shazeer可能是个时间旅行者。 我原本不相信他们的gated MLP （在写第3.3部分的时候，因为门控多层感知机让训练不稳定） ，但现在我信服了 （在添加了Canon 层之后，我们在第4.1部分对比了多层感知机和门控多层感知机） 。 正式认识一下，沙哥是谁？ 他是 Transformer八 ...

一日为谄媚者， 终身为谄媚者 作者｜冬梅、核子可乐 近日，OpenAI 在其官网发文称已回滚上周 ChatGPT 中的 GPT-4o 更新，目前用户使用的是行为更加平衡的早期版本。Altam 也在 X 上发帖说明了这一 调整。 为什会做这样的调整？因为最近不少用户发现 ChatGPT 越来越"谄媚"。 随着用户对于 ChatGPT "谄媚"行为的讨论越来越多，前微软高管、现 Spotify 首席技术官 Mikhail Parakhin 也发表了他对此事的看法。 Parakhin 认为，ChatGPT 并非一开始就以奉承用户为默认风格，不过由于用户对直接的人格反馈表现出强烈反感，OpenAI 决定调整聊天机器人，让其 更讨好用户。 Parakhin 表示："ChatGPT 的记忆功能首次推出时，初衷是让用户查看和编辑 AI 生成的个人资料。然而，即使是像"有自恋倾向"这样相对中性的表述， 也常常引发强烈反应。" "很快就发现人们敏感得可笑：'有自恋倾向'——'不，我没有！'，不得不隐藏它。因此才有了这批极度谄媚的 RLHF，"Parakhin 说道。 RLHF（基于人类反馈的强化学习）用于根据人们偏好的回应方式 ...