AI in Marketing

In March 2025, Anthropic CEO Dario Amodei predicted that AI could write 90% of code within three to six months, a claim that sparked widespread discussion and skepticism across the tech industry. This prediction highlighted the potential for AI to significantly impact software development, yet it also raised concerns about the feasibility and implications of such advancements. Current AI tools excel at specific tasks like autocompletion and generating short code snippets but fall short in handling complex, multi-module projects. Achieving 90% code generation would require AI systems to consistently produce production-grade outputs, including multi-file project synthesis, accurate API wiring, reliable testing, and secure-by-default patterns. However, many of these capabilities are still developing, and existing tools primarily assist with localized tasks rather than comprehensive system generation. Benchmarks indicate progress in token-level completion and unit test generation, but these do not equate to full-scale software delivery. Security remains a significant concern, with studies showing that a substantial portion of AI-generated code contains vulnerabilities. Enterprises adopting AI coding assistants report productivity gains in routine tasks but maintain human oversight for compliance and integration. Regulatory and standards considerations further complicate rapid adoption, especially in regulated sectors. While AI tools are becoming valuable aids in software development, they are not yet substitutes for human engineers. The prediction of AI writing 90% of code within a short timeframe appears optimistic, given current technological and governance constraints. For meaningful adoption, organizations should focus on pilot projects, invest in security and review processes, and develop skills in prompt engineering and model evaluation. The future of AI in coding will likely involve a balance of increased productivity and heightened security and governance measures.

Next up

Google has removed some AI-generated health summaries after The Guardian reported inaccuracies in the information provided for liver-related queries. These summaries, which lacked context and failed to consider factors like nationality and age, could potentially mislead patients, delaying necessary medical care. Google stated that while most AI Overviews are accurate, they are working to improve context where needed. This development occurs as AI's role in healthcare grows, with OpenAI reporting significant user engagement in health-related queries and launching ChatGPT Health. OpenAI also acquired the healthcare startup Torch, while Anthropic introduced AI tools for medical purposes. As AI integration in healthcare advances, ensuring accuracy and context remains crucial to avoid adverse patient outcomes.

And that brings me to

TechCrunch reports that Dr. Sina Bari, a surgeon and AI healthcare leader, has observed the potential dangers of using ChatGPT for medical advice, citing a case where a patient was misled by incorrect information from the AI. Despite these concerns, Dr. Bari supports the introduction of ChatGPT Health, a new version of the chatbot designed to provide private and secure health consultations without using user data for AI training. This development allows users to upload medical records and sync with health apps, raising privacy concerns among some experts. The widespread use of AI chatbots for health inquiries, with over 230 million users weekly, highlights the need for secure and accurate AI tools in healthcare. Industry professionals like Andrew Brackin see the potential for a more optimized version of ChatGPT for healthcare queries. However, AI hallucinations remain a challenge, with OpenAI's GPT-5 reportedly more prone to errors than other models. Dr. Nigam Shah from Stanford emphasizes the urgent need for accessible healthcare, suggesting AI could alleviate long wait times for medical consultations. He advocates for AI integration on the provider side to reduce administrative burdens, allowing doctors to focus more on patient care. Companies like Anthropic are also developing AI tools to streamline healthcare administration, aiming to save time and improve efficiency.

Which brings us to

Apple has announced a collaboration with Google to enhance its services, including a more personalized Siri, using Google's Gemini AI models. This partnership marks a shift from Apple's traditional approach of developing its own technology layers, reflecting its current reliance on Google's advanced AI capabilities. Analysts suggest this move is strategic for Apple, as it seeks to meet consumer demand for AI features, despite AI not being a primary factor in iPhone purchases. The deal, while potentially beneficial for users, may raise regulatory concerns due to the significant market power both companies hold. Previous collaborations between Apple and Google have been financially substantial, and this new partnership could attract scrutiny from regulators concerned about market dominance.

Moving on

Avalon GloboCare Corp. has regained compliance with Nasdaq's minimum stockholders’ equity requirement, allowing its stock to continue trading on the Nasdaq Capital Market. This development follows a notice from Nasdaq confirming the company's adherence to all listing standards, leading to the cancellation of a scheduled hearing. Avalon, a diversified company, focuses on precision diagnostic consumer products and intellectual property in cellular therapy and generative AI. The company is advancing its marketing efforts for a generative AI video platform and expanding its precision diagnostics and strategic assets. Avalon is also marketing the KetoAir™ breathalyzer, a Class I medical device registered with the FDA, and plans to explore additional diagnostic applications for this technology.

In other news

TechCrunch reports that Salesforce has launched a new AI-powered version of Slackbot, aiming for it to become as popular as OpenAI's ChatGPT. This updated Slackbot, available for Business+ and Enterprise+ customers, can perform tasks such as finding information, drafting emails, and scheduling meetings within Slack. It can also interact with other enterprise tools like Microsoft Teams and Google Drive, provided it has the necessary permissions. This integration allows users to work across various applications without leaving Slack. Salesforce and its competitors are heavily investing in AI to maintain and expand their market presence. The revamped Slackbot, first announced at Salesforce's Dreamforce conference, represents a significant shift from previous versions, retaining its name due to its established recognition. Salesforce CTO Parker Harris describes the new Slackbot as a "super agent" powered by generative AI, designed to enhance user experience. Internally, Slackbot has seen high adoption rates, indicating a strong product-market fit. Future plans for Slackbot include adding voice capabilities and internet browsing features, which Harris believes will benefit both Slack and Salesforce as a whole.

Shifting gears

The robotics industry is rapidly advancing, with significant developments expected by 2026. Key trends include the rise of predictive math, which will enable robots to anticipate and optimize tasks rather than just react. This advancement will allow robots to simulate scenarios quickly, enhancing their adaptability and efficiency. Additionally, imitation learning will transform robots from isolated units into cooperative teams that learn from each other and humans, improving workflow flexibility and resilience. Manufacturers are also shifting towards task-specific AI, integrating pre-trained systems for processes like welding and assembly, which will streamline operations and enhance quality. Furthermore, a new data economy is emerging, where anonymized data from robots will be shared with AI developers to improve applications, creating a cycle of continuous improvement. These innovations promise increased productivity, faster deployment, and reduced downtime, marking a significant evolution in robotics and physical AI. Anders Billesø Beck of Universal Robots emphasizes these trends as pivotal for the future of automation and human-robot collaboration.

Also in the news

Orsi Korman, an account director at Havas Red US, discusses the negative impact of jargon in business writing. Jargon, often found in corporate language and AI-generated drafts, can dilute meaning, erode trust, and make writing difficult to understand. Korman identifies seven commonly overused words: deliver, impact, alignment, leverage, optimize, unlock, and synergy. These words often obscure clarity and should be replaced with more specific language. For example, instead of saying "deliver a best-in-class employee experience," one could say "create a more collaborative and supportive employee experience." Korman advises replacing abstract terms with specific ones, using conversational verbs, and reading text aloud to ensure clarity. Clear communication is essential for effective writing.