At TechCrunch Disrupt 2024, Ashton Kutcher, co-founder of Sound Ventures, shared his belief that every company will eventually incorporate AI, though he doubts there will be a single “winner” in the space. Kutcher emphasised the transformative potential of foundational AI models, which he views as essential to future innovation across industries. “There will not be a company in the world that is not, in some way, using AI,” he noted, adding that foundational companies in AI could become some of the most valuable in history.
Kutcher, alongside Sound co-founders Guy Oseary and Effie Epstein, explained that Sound Ventures is betting heavily on AI, with a $265 million fund backing major AI firms like OpenAI and Anthropic. Kutcher also shared that OpenAI’s CEO, Sam Altman, supported Sound’s multi-company AI investments, with the firm carefully maintaining confidentiality across its portfolio.
While some are apprehensive about AI’s rapid growth, Kutcher compared the technology’s potential impact to transformative past innovations like personal computers and cars. Advising founders, he highlighted the importance of strong teams over polished pitch decks, noting that real value lies in people, market insight, and breakthrough ideas.
SoftBank CEO Masayoshi Son has reaffirmed his belief that artificial super intelligence (ASI) will become a reality by 2035. Speaking at a conference in Riyadh, Saudi Arabia, Son explained that ASI would be 10,000 times smarter than the human brain. He highlighted the need for hundreds of billions of dollars in investment to achieve this groundbreaking development.
Son mentioned plans to save funds for his next major venture but did not provide specifics. He estimates that generative AI will require $900 trillion in future capital expenditure, especially for data centres and chip production. He also suggested that Nvidia is undervalued, given the vital role its chips will play in this industry.
Although Son has been a strong proponent of new technology, his investment track record has been mixed. The Vision Fund, launched in 2017, reshaped venture capital, but many of its high-growth startups have since lost value, contributing to a $2.4 billion loss as of June 2024.
Saudi Arabia’s Public Investment Fund (PIF) had invested $45 billion in the first Vision Fund. Despite recent losses, Son remains optimistic about the future of AI and its potential to transform industries worldwide.
Meta is working on a new AI search engine to lessen its reliance on Google and Microsoft’s Bing. The move places Meta among other tech giants, such as OpenAI, Google, and Microsoft, in the race to dominate the evolving AI-powered search landscape.
The new search tool aims to enhance Meta’s chatbot on WhatsApp, Instagram, and Facebook by offering conversational responses to real-time queries about news and events. Meta currently depends on Google and Bing to provide users with information on topics like news, stock markets, and sports.
As competition intensifies, Google is pushing its Gemini AI model into core services, including Search, to offer more interactive and intuitive experiences. OpenAI, meanwhile, continues to use Bing, leveraging its close partnership with Microsoft for topical queries.
The use of web data to train AI systems and build search engines has sparked debates about copyright and fair compensation. Meta recently announced that its chatbot would incorporate Reuters content to provide up-to-date answers to questions related to news and current events.
Sierra, a young AI software startup co-founded by former Salesforce co-CEO Bret Taylor, has secured $175 million in new funding led by Greenoaks Capital. This latest round gives the company a valuation of $4.5 billion, a significant jump from its earlier valuation of nearly $1 billion. Investors such as Thrive Capital, Iconiq, Sequoia, and Benchmark have also backed the firm.
Founded just a year ago, Sierra has already crossed $20 million in annualised revenue, focusing on selling AI-powered customer service chatbots to enterprises. It works with major clients, including WeightWatchers and Sirius XM. The company claims its technology reduces ‘hallucinations’ in large language models, ensuring reliable AI interactions for businesses.
The rising valuation reflects investor enthusiasm for applications in AI that generate steady revenue, shifting from expensive foundational models to enterprise solutions. Sierra operates in a competitive space, facing rivals such as Salesforce and Forethought, but aims to stand out through more dependable AI performance.
Bret Taylor, who also chairs OpenAI’s board, co-founded Sierra alongside former Google executive Clay Bavor. Taylor previously held leadership roles at Salesforce and oversaw Twitter’s board during its takeover by Elon Musk. Bavor, who joined Google in 2005, played key roles managing Gmail and Google Drive.
On 21 and 24 October, DiploFoundation provided just-in time reporting from the UN Security Council sessions on scientific development and on women, peace, and security. Supported by Switzerland, this initiative aims to enhance the work of the UN Security Council and the broader UN system.
At the core of this effort is DiploAI, an advanced platform shaped by years of training on UN materials, which played a crucial role in unlocking the knowledge generated by the Security Council’s deliberations. This knowledge, often trapped in video recordings and transcripts, is now more accessible, providing valuable insights for diplomacy and global peace.
Unlocking the power of AI for peace and security
AI-supported reporting from the UN Security Council (UNSC) demonstrates the potential of combining cutting-edge technology with deep expertise in peace and security. This effort is part of ongoing work by DiploAI, which has been providing detailed reports on Security Council sessions in 2023-2024 and has covered the UN General Assembly (UNGA) for eight consecutive years. DiploAI is actively contributing to expanding the UN’s knowledge ecosystem.
Seamless interplay between experts and AI
The success of this initiative lies in the seamless interplay between DiploAI and security experts well-versed in UNSC procedures. The collaboration began with tailoring the AI system to the unique needs of the Council, using input from experts and diplomats to build a relevant knowledge base. Experts supplied key documents and session materials, which enhanced the AI’s contextual understanding. Feedback loops on keywords, topics, and focus areas ensured the AI’s output remained both accurate and diplomatically relevant.
A pivotal moment in this collaboration was the analysis of New Agenda for Peace , where Security Council experts helped DiploAI identify over 400 critical topics, laying the foundation for a comprehensive taxonomy on peace and security at the UN. This expertise, combined with DiploAI’s technical capabilities, has resulted in an AI system attuned to the subtleties of diplomatic language and priorities. Furthermore, the project introduced a Knowledge Graph—a visual tool for displaying sentiment and relational analysis between statements and topics—which adds new depth to the analysis of Council sessions.
Building on this foundation, DiploAI developed a custom chatbot capable of moving beyond standard Q&A interactions. By integrating data from all 2024 sessions and associated documents, the chatbot allows users to interact conversationally with the content, providing in-depth answers and real-time insights. This evolution marks a significant leap forward in accessing and understanding diplomatic data—shifting from static reports to interactive exploration of session materials.
AI and diplomatic sensitivities
The development of DiploAI’s Q&A module, refined through approximately ten iterations with feedback from UNSC experts, underscores the value of human-AI(-nism) collaboration. This module addresses essential diplomatic questions, with iterative refinements ensuring that responses meet the Council’s standards for accuracy and relevance. The result is an AI system capable of addressing critical inquiries while respecting the sensitivity required in diplomatic settings.
What’s new?
DiploAI’s suite of tools—including real-time meeting transcription and analysis—has transformed reporting and transparency at the UNSC. By integrating customized AI systems like retrieval-augmented generation (RAG) and knowledge graphs, DiploAI adds context, depth, and relevance to the extracted information. Trained on a vast corpus of diplomatic knowledge generated at Diplo over the last two decades, the AI system generates context-specific responses, providing comprehensive answers to questions about transcribed sessions.
Here are some numbers from 10 UNSC meetings that took place between January 2023 and October 2024:
Number of speakers and speech length
Unique speakers: 185
Total time: 201,221.25 min – 2.0 days, 7.0 hours, 53.0 minutes, 41.0 seconds
Total speeches: 583
Total length: 396,172 words, or 0.67 ‘War and Peace’ books
Frequency of selected topics
Name of the topic
Number of times it was mentioned
Name of the session
development
1,665 mentions
The session that most mentioned development: UNSC meeting: Peace and common development (919 mentions)
climate change
451 mentions
The session that most mentioned climate change: UNSC meeting: Climate change and food insecurity (329 mentions)
human rights
360 mentions
The session that most mentioned human rights: UNSC meeting: Peace and common development (93 mentions)
civilians
136 mentions
The session that most mentioned civilians: UNSC meeting: Peacekeeping (72 mentions)
international humanitarian law
27 mentions
The session that most mentioned international humanitarian law: UNSC meeting: Multilateral cooperation (6 mentions)
In conclusion…
DiploAI’s reporting from the Security Council, supported by Switzerland, shows how AI can enhance diplomacy while staying grounded in human expertise and practical needs. This blend of technical capability and domain-specific knowledge demonstrates how AI, when developed collaboratively, can contribute to more inclusive, informed, and impactful diplomacy.
An AI transcription tool called Whisper, developed by OpenAI and used by thousands of clinicians and health systems, has come under scrutiny after researchers found it sometimes produces inaccurate transcriptions. Whisper, which powers the medical transcription tool from the company Nabla, has reportedly transcribed around 7 million medical conversations. While it accurately summarises many doctor-patient exchanges, researchers from Cornell University and the University of Washington discovered instances where the AI-generated entirely fabricated sentences, sometimes even adding irrelevant or nonsensical phrases.
The study, which was presented at the Association for Computing Machinery FAccT conference in Brazil in June, highlighted that Whisper made errors in about 1 percent of transcriptions, often producing ‘hallucinations’ — fabricated statements in response to silences during conversations. These inaccuracies were especially common in audio samples featuring patients with aphasia, a language disorder that results in frequent pauses. In one case, Whisper inserted phrases that were more typical of a YouTube video, such as “Thank you for watching!”
Nabla, aware of the issue, has stated it is working on solutions to mitigate these hallucinations. In response, OpenAI emphasised its commitment to reducing such errors, particularly in high-stakes situations like healthcare. An OpenAI spokesperson noted that Whisper’s usage policies discourage its application in critical decision-making contexts and that guidance for open-source use advises against deployment in high-risk domains.
The study’s findings underscore the complexities of applying AI tools in sensitive settings like healthcare, where precise communication is vital. With Whisper being used across 40 healthcare systems, the issue raises broader questions around the suitability of AI transcription tools in medical environments and the ongoing need for oversight in their deployment.
Google has extended its AI Overviews in Search to more than 100 countries and territories. Initially launched in the US in May, the feature provides summarised snapshots at the top of search results. It now serves over one billion users globally each month.
The expanded rollout introduces more language options, including English, Hindi, Indonesian, Japanese, Portuguese, and Spanish. Google aims to enhance the usability of the tool, offering new features like in-line links, which improve website traffic by embedding source links directly within the text.
AI Overviews are also playing a role in the company’s advertising strategy. Ads will now appear within the AI-generated summaries for mobile users in the US, marking a new direction for Google’s ad business by integrating advertising more seamlessly.
Despite some challenges at launch, including incorrect information that raised concerns, Google has made significant improvements. Fine-tuning efforts are ongoing, and the feature has also been introduced to Google Shopping, further expanding its presence across the platform.
A new podcast titled Virtually Parkinsonbrings back the voice of Sir Michael Parkinson, using AI technology to simulate the late chat show host. Produced by Deep Fusion Films with support from Parkinson’s family, the series aims to recreate his interview style across eight episodes, featuring new conversations with prominent guests.
Mike Parkinson, son of the late broadcaster, explained that the family wanted listeners to know the voice is an AI creation, ensuring transparency. He noted the project was inspired by conversations he had with his father before he passed, saying Sir Michael would have found the concept intriguing, despite being a technophobe.
The release comes amid growing controversy around AI’s role in the creative arts, with many actors and presenters fearing it could undermine their careers. Though AI is often criticised for replacing real talent, Parkinson’s son argued that the podcast offers a unique way to extend his father’s legacy, without replacing a living presenter.
Co-creator Jamie Anderson clarified that the AI version acts as an autonomous host, conducting interviews in a way reflective of Sir Michael’s original style. The podcast seeks to introduce his legacy to younger audiences, while also raising ethical questions about the use of AI to recreate deceased individuals.
The Open Source Initiative (OSI) has introduced version 1.0 of its Open Source AI Definition (OSAID), setting new standards for AI transparency and accessibility. Developed over the years in collaboration with academia and industry, the OSAID aims to establish clear criteria for what qualifies as open-source AI. The OSI says the definition will help align policymakers, developers, and industry leaders on a common understanding of ‘open source’ in the rapidly evolving field of AI.
According to OSI Executive Vice President Stefano Maffulli, the goal is to make sure AI models labelled as open source provide enough detail for others to recreate them and disclose essential information about training data, such as its origin and processing methods. The OSAID also emphasises that open source AI should grant users freedom to modify and build upon the models, without restrictive permissions. While OSI lacks enforcement power, it plans to advocate for its definition as the AI community’s reference point, aiming to combat “open source” claims that don’t meet OSAID standards.
The new definition comes as some companies, including Meta and Stability AI, use the open-source label without fully meeting transparency requirements. Meta, a financial supporter of the OSI, has voiced reservations about the OSAID, citing the need for protective restrictions around its Llama models. In contrast, OSI contends that AI models should be openly accessible to allow for a truly open-source AI ecosystem, rather than restricted by proprietary data and usage limitations.
Maffulli acknowledges the OSAID may need frequent updates as technology and regulations evolve. OSI has created a committee to monitor its application and adjust as necessary, with an eye on refining the open-source definition to address emerging issues like copyright and proprietary data.
Alphabet’s recent earnings report showcases a robust 35% surge in Google Cloud revenue for the third quarter, signalling a strong demand for AI-driven cloud computing and setting a promising tone for competitors Microsoft and Amazon. The jump, the fastest pace of growth in two years, has sent Alphabet’s shares up by 5.5%, reflecting investor optimism about the company’s expanding foothold in the cloud sector. Analysts had anticipated growth of around 29%, making this an even stronger-than-expected outcome for Google’s cloud division.
While Alphabet’s cloud segment remains smaller than Amazon’s AWS and Microsoft’s Azure, its performance has been noteworthy, accounting for 13% of Alphabet’s third-quarter revenue, up from 11% a year prior. This consistent growth aligns with Google’s continued investment in AI-powered tools and custom chips, like the Tensor Processing Unit (TPU), which have set it apart from competitors by enhancing cloud capabilities for AI applications. Analysts like Angelo Zino of CFRA Research believe Google may deliver the most impressive cloud growth numbers this quarter.
Google has been pouring resources into AI advancements across its cloud and search businesses, including its chatbot Gemini, which offers AI-generated code, data processing, and cybersecurity tools. These innovations, combined with the Vertex AI platform that enables custom model creation, have made Google Cloud a compelling choice for organisations seeking advanced AI solutions.
Investments in AI are also fueling Alphabet’s core advertising business, which rose 10% in the quarter, and supporting its broader cloud infrastructure expansion. Google has announced plans to invest billions in new data centres worldwide to strengthen its market presence and technological capabilities.
In its latest earnings call, Alphabet’s new CFO, Anat Ashkenazi, confirmed that capital expenditures in 2025 are expected to surpass this year’s, underlining the company’s commitment to scaling its AI and cloud capabilities. Analysts like Gil Luria of D.A. Davidson emphasise that Google Cloud’s outstanding performance in recent quarters showcases Alphabet’s success in turning AI investments into substantial revenue growth, solidifying its position in a competitive market.
