Birth order is a fascinating field of observation. I love to guess whether someone is an eldest, middle, or youngest sibling based on their behavior as an adult. While I’m a firstborn who fits a lot of conventional birth-order wisdom, I’d select “choose your own adventure” youngest-sibling energy to describe today’s B2B customer marketers. Exciting choices and blazed trails abound, as the customer marketing role has expanded well beyond cross-sell and upsell to encompass a vast set of responsibilities including adoption, digital engagement, and customer advocacy. Forrester’s new State Of Customer Engagement Survey, 2024, shows just how expansive this role has become. Variety is thrilling. But I’d be remiss as the eldest sib if I didn’t give the same warning to customer marketers that I give to my own youngest child: Be ambitious but not careless. Long-term success requires you to take on the right things, not everything. Ambition And Recognition Are Worthy Goals; Scope Creep Is Not Our new data overview report on B2B Customer Marketing Responsibilities shows adoption, reference management, and digital programs among the top three customer marketing remits. The potential pitfall of that range of responsibilities: A whopping 86% of customer marketers said that their team has “too many competing priorities.” If that feels familiar, you might be spreading yourself too thin and creating unnecessary friction with other functions. To avoid this, you must prioritize. Here are some ideas: Write it down. A customer marketing charter captures customer marketing’s mission statement, key initiatives, stakeholders, and most important metrics. From experience with our customer marketing clients, I promise it’s not the “homework” it might feel like. A clear charter inspires your team by reminding them what customer marketing is about, helps socialize your contribution to the rest of the company, and protects against inevitable scope creep. We’ll have a customer marketing charter workshop at our B2B Summit in early 2025 to get you started on your own charter. Keep your friends close. Successful companies protect against overlap and border skirmishes by ensuring that post-sale engagement functions collaborate rather than override each other. As a customer marketer, you should work with customer success to support adoption and usage programs. You should partner with sales to make sure that reference programs align with buyers’ needs. You should work with portfolio marketing and product teams to build resonant digital experiences such as online communities and customer portals and glean valuable insight from customer interactions. Keep customer value front and center. The simplest litmus test? Repeatedly ask yourself whether a program or tactic contributes to maximizing value for the company and the customer. That might be economic or functional value. It might also, or instead, appear as symbolic or experiential value. Be willing to evolve. Maybe something worked in the past, but does it work still? If not, be willing to rebalance, refocus, and deploy your efforts where they will do the most measurable good. Read more about how customer marketers spend their time and resources, and where they’d like to do more, in the data overview report, B2B Customer Marketing Responsibilities. Contact us if you’d like to workshop your own customer marketing charter. source

By Christopher Cole ( December 4, 2024, 6:12 PM EST) — Frontier Communications has agreed to pay almost $288,000 to end a Federal Communications Commission probe into a Wisconsin agency’s claims that the internet service provider submitted inaccurate information to the FCC during a challenge to data used in mapping national broadband service, according to a consent decree made public Wednesday…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More The recent release of OpenAI o1 has brought great attention to large reasoning models (LRMs), and is inspiring new models aimed at solving complex problems classic language models often struggle with. Building on the success of o1 and the concept of LRMs, researchers at Alibaba have introduced Marco-o1, which enhances reasoning capabilities and tackles problems with open-ended solutions where clear standards and quantifiable rewards are absent. OpenAI o1 uses “inference-time scaling” to improve the model’s reasoning ability by giving it “time to think.” Basically, the model uses more compute cycles during inference to generate more tokens and review its responses, which improves its performance on tasks that require reasoning. o1 is renowned for its impressive reasoning capabilities, especially in tasks with standard answers such as mathematics, physics and coding.  However, many applications involve open-ended problems that lack clear solutions and quantifiable rewards. “We aimed to push the boundaries of LLMs even further, enhancing their reasoning abilities to tackle complex, real-world challenges,” Alibaba researchers write. Marco-o1 is a fine-tuned version of Alibaba’s Qwen2-7B-Instruct that integrates advanced techniques such as chain-of-thought (CoT) fine-tuning, Monte Carlo Tree Search (MCTS) and reasoning action strategies. The researchers trained Marco-o1 on a combination of datasets, including the Open-O1 CoT dataset; the Marco-o1 CoT dataset, a synthetic dataset generated using MCTS; and the Marco-o1 Instruction dataset, a collection of custom instruction-following data for reasoning tasks. Marco-o1 uses CoT and MCTS to reason about tasks (source: arXiv) MCTS is a search algorithm that has proven to be effective in complex problem-solving scenarios. It intelligently explores different solution paths by repeatedly sampling possibilities, simulating outcomes and gradually building a decision tree. It has proven to be very effective in complex AI problems, such as beating the game Go. Marco-o1 leverages MCTS to explore multiple reasoning paths as it generates response tokens. The model uses the confidence scores of candidate response tokens to build its decision tree and explore different branches. This enables the model to consider a wider range of possibilities and arrive at more informed and nuanced conclusions, especially in scenarios with open-ended solutions. The researchers also introduced a flexible reasoning action strategy that allows them to adjust the granularity of MCTS steps by defining the number of tokens generated at each node in the tree. This provides a tradeoff between accuracy and computational cost, giving users the flexibility to balance performance and efficiency. Another key innovation in Marco-o1 is the introduction of a reflection mechanism. During the reasoning process, the model periodically prompts itself with the phrase, “Wait! Maybe I made some mistakes! I need to rethink from scratch.” This causes the model to re-evaluate its reasoning steps, identify potential errors and refine its thought process. “This approach allows the model to act as its own critic, identifying potential errors in its reasoning,” the researchers write. “By explicitly prompting the model to question its initial conclusions, we encourage it to re-express and refine its thought process.” To evaluate the performance of Marco-o1, the researchers conducted experiments on several tasks, including the MGSM benchmark, a dataset for multi-lingual grade school math problems. Marco-o1 significantly outperformed the base Qwen2-7B model, particularly when the MCTS component was adjusted for single-token granularity.  Different versions of Marco-o1 vs base model (source: arXiv) However, the primary objective of Marco-o1 was to address the challenges of reasoning in open-ended scenarios. To this end, the researchers tested the model on translating colloquial and slang expressions, a task that requires understanding subtle nuances of language, culture and context. The experiments showed that Marco-o1 was able to capture and translate these expressions more effectively than traditional translation tools. For instance, the model correctly translated a colloquial expression in Chinese, which literally means, “This shoe offers a stepping-on-poop sensation”, into the English equivalent, “This shoe has a comfortable sole.” The reasoning chain of the model shows how it evaluates different potential meanings and arrives at the correct translation. This paradigm can prove to be useful for tasks such as product design and strategy, which require deep and contextual understanding and do not have well-defined benchmarks and metrics. Example of reasoning chain for translation task (source: arXiv) A new wave of reasoning models Since the release of o1, AI labs are racing to release reasoning models. Last week, Chinese AI lab DeepSeek released R1-Lite-Preview, its o1 competitor, which is currently only available through the company’s online chat interface. R1-Lite-Preview reportedly beats o1 on several key benchmarks. The open source community is also catching up with the private model market, releasing models and datasets that take advantage of inference-time scaling laws. The Alibaba team released Marco-o1 on Hugging Face along with a partial reasoning dataset that researchers can use to train their own reasoning models. Another recently released model is LLaVA-o1, developed by researchers from multiple universities in China, which brings the inference-time reasoning paradigm to open-source vision language models (VLMs).  The release of these models comes amidst uncertainty about the future of model scaling laws. Various reports indicate that the returns on training larger models are diminishing and might be hitting a wall. But what’s for certain is that we are just beginning to explore the possibilities of inference-time scaling. source

LinkedIn Email Facebook Twitter WhatsApp The post 行業研究報告: 2024 年人工智慧新創產業獲獎者 appeared first on VeriMedia. source

By Bryan Koenig ( December 3, 2024, 8:44 PM EST) — The U.S. Department of Justice and Federal Trade Commission drew criticism Tuesday from former officials who targeted the agencies’ stance on Google’s Ninth Circuit fight against the mandated opening of the Android Play Store, with the officials warning in an amicus brief against “compulsory sharing obligations.”… Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI appears to be launching a new subscription tier offering for its signature chatbot product, ChatGPT. Screenshots posted on X by third-party AI engineer Tibor Blaho show the new tier, ChatGPT Pro, priced at $200 U.S. dollars per month, 10x the amount currently charged for the ChatGPT Plus individual offering. It is also the highest sum compared to ChatGPT Team ($30), Enterprise (varies but estimated to be $60-$100), and Edu (varies but estimated at $12 per month). Yet the Pro plan will grant users access to “the best of OpenAI with the highest level of access” according to the screenshot, which includes “unlimited” access to its newest o1 and o1-mini reasoning models and even “more compute,” a.k.a. graphics-processing unit (GPU) capacity that OpenAI has for serving up model inferences (live models users can interact with). Plus, meanwhile, only grants users “limited access” to the o1 and o1-mini reasoning models. VentureBeat has reached out to OpenAI contacts for confirmation or further information on ChatGPT Pro and will update when we hear back. The news comes just a day after OpenAI said it would today begin 12 days of holiday-themed announcements entitled “12 Days of OpenAI,” an obvious allusion to the “12 Days of Christmas” song and tradition. OpenAI is expected to make its announcement today in an hour, around 10 am PT. source

Social media is vast, with several distinct sites and innumerable accounts spread across them. Because of this complexity, researchers cast a wide net across the most popular sites to create an inclusive list of influencers who discuss news. First, we developed a set of keywords related to current events and civic issues across 45 topics that were relevant to broad audiences in the U.S. in early 2024. Then we used a set of social media marketing tools to find social media accounts with over 100,000 followers that used those keywords in early 2024. While the keywords used were focused on issues important to the United States, we could not verify the location of the individuals running these accounts. One thing quickly became clear: Discussion of news, politics and civic issues is widespread on social media. This initial screening led to a list of more than 28,000 accounts across Facebook, Instagram, TikTok, X and YouTube. Many of these 28,000 accounts are not focused on news most of the time. Researchers examined recent posts from each account to determine if they were talking about news, politics and civic issues regularly, and not just rarely. We also verified that accounts were run by individuals and not organizations. Just over 2,500 accounts – and 2,058 influencers, since influencers can have an account on multiple sites – met these criteria. To create a more manageable group to analyze, researchers then sampled 500 news influencers from this set of 2,058. These 500 form the set of popular news influencers who are examined in this report. Our analysis includes: Detailed human coding of all 500 news influencers to determine where they post and who they are. This includes a look at other places they may post beyond the five major sites we searched for influencers on. Human-validated machine coding of over 100,000 posts from these news influencers during three weeks in July and August 2024, to determine what these accounts are posting about. For more information, refer to the methodology. Where are news influencers found? While news influencers can be found on a wide range of social media sites, they are concentrated on a few. The vast majority of news influencers studied (85%) have an account on X, formerly known as Twitter. Half are on Instagram, while a slightly smaller share (44%) are on YouTube. Roughly three-in-ten are on Facebook (32%), Threads (30%) and TikTok (27%). And smaller shares are on several other sites, including 5% who are on Donald Trump’s site, Truth Social. (A 2022 study found that most prominent accounts on Truth Social were individuals.) More Americans say they regularly get news (in general, not just from news influencers) from Facebook (33%) and YouTube (32%) than other social media sites, while smaller shares get news on Instagram (20%), TikTok (17%) and X (12%). At the same time, a much larger share of X users than users of other sites say getting news is a reason they use the site. Altogether, about two-thirds of news influencers are on more than one social media site, including 27% who are on five or more sites. Meanwhile, 34% are on just one site – most often X. Other ways news influencers connect with their audiences We also looked at other ways that news influencers connect with their audiences beyond social media. About a third (34%) host a podcast, and nearly a quarter (22%) send an email newsletter to subscribers. A small share of news influencers in the sample (6%) have an official Discord server, a chat site on which they can often have private, ongoing chats with fans. Most news influencers (59%) also make direct financial appeals to their followers. About half (49%) offer a paid subscription to their content, 29% solicit donations, and 21% sell branded merchandise like stickers, coffee mugs and apparel. source

I recently had the opportunity to write a research report with Vicki Brown to tackle questions that we often get from clients: What is the difference between a buyer’s journey map and a revenue waterfall, and do they work together? The Buyer’s Journey Map Buyer’s journey maps are developed to represent the buyer’s view of the purchasing process. They help us understand what information buyers seek, when they need it, and where they go to find it. With that information, B2B marketers can build better go-to-market strategies and engagement plans. The B2B Revenue Waterfall™ On the contrary, the B2B Revenue Waterfall focuses on internal processes, tracking targeted opportunities as they move through the waterfall stages. The goal is for an organization to measure the flow of demand, inform demand program planning to increase the volume of opportunities, and improve the velocity of existing opportunities. Trying to conflate the two is dangerous and hinders the purpose of each framework. It also harms both buyers and sellers, because the waterfall stage for the group may not always equal where every buyer is in their journey. Do They Work Together? The answer: sometimes. Insights from both Forrester’s B2B Buyer’s Journey Map Framework and the B2B Revenue Waterfall can inform how to improve the other, but they are ultimately designed to do two different things. How does your organization plan (external view) and manage (internal view) demand generation programs in a way that serves both the buyer’s needs and the organization’s need to measure progress and manage resources? Forrester clients: Let’s chat more via a Forrester guidance session. Forrester clients also can access our report here. source