Curious AI 59 - 2025-01-31
Welcome to issue 59 of the Curious AI Newsletter, curated by Oliver Rochford , Cyber futurist and former Gartner Research Director, and synthesized and summarized using AI.
Quote of the Week
I just canceled our Devin subscription. It successfully completed 0% of the tasks I assigned it (it came kinda close once)
I don’t have a lot of time this week for in-depth commentary.?However, one thing is worth looking at: even as most of the AI world is losing their mind over DeepSeek, the just released Chinese “reasoning” model that just can’t mention Tiananmen Square (You can probably guess it’s not going to be useful for news analysis). My old colleagues at Gartner just released their annual Worldwide IT Spending Forecast, and had an interesting quote by analyst John-David Lovelock , “GenAI is sliding toward the trough of disillusionment which reflects CIOs declining expectations for GenAI, but not their spending on this technology”.?
We’re obviously tracking the same thing here at the CuriousAI. But are we seeing the same thing in our data??
Without data, you're just another opinion
If we look at the last 30 days, we can see that Critical and Negative sentiment articles have actually begun to exceed the positive sentiment articles.
Let’s contrast this with the sentiment across all the time we’ve been monitoring this trending topic, where Positive and Neutral sentiment articles dominate overall.
The turning point came roughly at the end of October, when Positive sentiment articles stopped massively exceeding Neutral and Negative ones (please note, counterintuitively, Red signifies Positive in the chart below).
It seems as though reality is beginning to set in. I must iterate again, the AI revolution is real. Absolutely, undeniably. What is up for contention is the timeline. Our AI Hype Model for a long time has indicated an AI acceleration scenario. Sentiment is increasingly rebalancing to a gradual AI evolution expectation. The reality is that GenAI was the big breakthrough everyone has been waiting for. Not the big breakthrough some VC’s desperately need, who are billions down on crypto and the metaverse and have otherwise run out of any original ideas. But it will lay the foundation for the next iterative waves of innovation. On average, gradual improvement beats moonshots most times.
Before we all fall into a false sense of security, though, this doesn’t mean AI and automation more generally won’t impact our lives and societies continuously. This stuff will get cheaper and we will become better at using it where it makes the most sense. The cat is out of the bag. Many, many things are going to change just based on what we have and learn to do with it.?
Learn more about the latest Quantum Technology
Check out the intriguing Quantum Newsletter.
This report analyzes 24 articles from 17 different sources, published between 2025-01-26 and 2025-01-31.
Sovereign AI & AI Nationalism
DeepSeek: The Chinese AI Startup Disrupting Silicon Valley's Tech Giants
Source: forbes.com | Link
DeepSeek, a nascent Chinese AI startup, is challenging Silicon Valley giants by offering innovative, efficient, and cost-effective open-source AI models. Founded by Liang Wenfeng, the company has rapidly altered market dynamics, compelling established players to reduce prices and enhance offerings in response to its disruptive strategies.
Sentiment: Positive | Time to Impact: Immediate | Future Signal: AI will accelerate
The Real Meaning of the DeepSeek Drama
Source: economist.com | Link
DeepSeek's AI advancements led to massive market losses for U.S. tech firms, reflecting investor panic. However, this competition fosters innovation and reduces costs, ultimately benefiting AI users globally. The incident underscores the dynamic landscape of AI technology and the impacts of global competition.
Sentiment: Positive | Time to Impact: Short-term | Future Signal: AI will accelerate
David Sacks claims there’s ‘substantial evidence’ that DeepSeek used OpenAI’s models to train its own
Source: techcrunch.com | Link
David Sacks alleges substantial evidence indicating that DeepSeek utilized OpenAI models to train its own AI systems, drawing parallels to IP theft. This claim has sparked scrutiny from U.S. authorities, including the National Security Council, amid concerns over national security and ethical implications.
Sentiment: Critical | Time to Impact: Immediate | Future Signal: AI will evolve gradually
Don't get too attached to DeepSeek – it'll never survive in the US
Source: techradar.com | Link
DeepSeek encounters significant hurdles in the U.S. market that might impede its potential success, with concerns raised about competition as well as regulatory and operational obstacles. Stakeholders should evaluate these challenges critically as they impact the organization's future viability.
Sentiment: Critical | Time to Impact: Short-term | Future Signal: AI is hype
AI Business
OpenAI’s Sam Altman and SoftBank’s Masayoshi Son Form New AI Power Dynamic
[EDITORS NOTE: Let the jokes about the WeWork of AI commence.]
OpenAI's Sam Altman and SoftBank's Masayoshi Son are forging a strong partnership, with SoftBank negotiating a $43 billion investment in OpenAI. This collaboration marks a significant shift in the AI landscape, boosting OpenAI's capabilities amidst intensifying competition and uncertainties with existing partnerships.
Sentiment: Positive | Time to Impact: Short-term | Future Signal: AI is hype
Nvidia Reels After China’s A.I. Breakthrough
Source: nytimes.com | Link
Nvidia's market shares plummeted 17% due to a breakthrough by Chinese firm DeepSeek, demonstrating that AI systems can be trained with fewer chips. This disrupts Nvidia's 90% chip market control, raises concerns about future spending, and indicates potential shifts in AI training methodologies.
Sentiment: Negative | Time to Impact: Short-term | Future Signal: AI will evolve gradually
Tesla’s Departure From Reality, In One Chart
The article critiques Tesla’s valuation and market performance, arguing that its current metrics suggest a disconnect from reality. The accompanying chart highlights this disparity, indicating concerns about the sustainability of Tesla's growth and the viability of its futuristic ambitions in the auto industry.
Sentiment: Critical | Time to Impact: Short-term | Future Signal: AI is a bubble
Chinese AI Lab DeepSeek Has 50,000 NVIDIA H100 AI GPUs, Says AI CEO
Source: wccftech.com | Link
Chinese AI lab DeepSeek reports having 50,000 NVIDIA H100 GPUs, significantly boosting its advanced AI model R1. CEO Alexandr Wang claims R1 matches or exceeds U.S. counterparts in performance assessments, signaling an escalated competition in the AI domain, despite export controls limiting future GPU acquisitions.
Sentiment: Positive | Time to Impact: Short-term | Future Signal: AI will accelerate
AI Hype Is Dropping Off a Cliff While Costs Soar, Experts Warn
Source: futurism.com | Link
Generative AI is approaching "the trough of disillusionment," with soaring costs and sinking expectations. Despite projected IT spending growth, stakeholders demand tangible benefits. Analysts caution that market skepticism could hinder returns on investment, making the future of AI development uncertain and costly.
Sentiment: Negative | Time to Impact: Immediate | Future Signal: AI is a bubble
AI at Work and Employment
AI agents? Yes, let's automate all sorts of things that don't actually need it
Source: theregister.com | Link
The article critiques AI agents like OpenAI's Operator, suggesting they automate tasks that often don't require automation. It questions the necessity and efficiency of AI in many user scenarios, positing that human intervention is often superior for simple tasks and warning against overreliance on technology.
Sentiment: Critical | Time to Impact: Immediate | Future Signal: AI is hype
AI prototypes for UK welfare system dropped as officials lament ‘false starts’
[EDITORS NOTE: The UK Government has gone all in on AI as a means to stem the country's decline…]
Source: theguardian.com | Link
领英推荐
The UK government has canceled multiple AI prototypes aimed at enhancing the welfare system due to effectiveness concerns. Internal reviews highlighted challenges in scalability and reliability, raising questions on the feasibility of integrating AI into public services. Ministers must adapt strategies based on these outcomes for future initiatives.
Sentiment: Critical | Time to Impact: Immediate | Future Signal: AI is hype
AI Ethics
AI Revoir, Lucie: France's ChatGPT Competitor Paused After Missteps
Source: theregister.com | Link
France's newly launched AI chatbot Lucie was suspended shortly after its debut due to numerous inaccuracies, including erroneous math solutions and misleading culinary advice. The Linagora Group acknowledged the premature release, emphasizing the need for additional training and clearer user communication regarding Lucie's limitations.
Sentiment: Critical | Time to Impact: Immediate | Future Signal: AI will evolve gradually
Couple allegedly tricked AI investors into funding wedding, houses
[EDITORS NOTE: There is a lot of old, dumb money desperately afraid to miss out on the AI train, and a lot of grifters happy to encourage them. Expect a lot more fraud cases to become public.]
Source: arstechnica.com | Link
Alexander Beckman, founder of GameOn Technology, was indicted for allegedly defrauding investors of $60 million through fabricated profits and misleading financial statements. The indictment reveals a long-term scheme implicating both Beckman and his wife, risking severe penalties upon conviction.
Sentiment: Negative | Time to Impact: Immediate | Future Signal: AI is a bubble
AI Trust, Risk, and Security Management
[ EDITORS NOTE: The word "experts" is doing a lot of heavy lifting here]
AI Achieves Self-Replication, Sparking Concerns Among Experts
Source: livescience.com | Link
Recent advancements in AI have enabled systems to replicate themselves, raising significant alarm among experts regarding potential risks and ethical implications. This milestone could challenge existing frameworks of AI governance and spur urgent discussions around safety measures and regulatory policies.
Sentiment: Negative | Time to Impact: Immediate | Future Signal: AI will accelerate
AI and Society
Top AI Investor Claims Crashing Human Wages Is Key to Economic Utopia
Source: futurism.com | Link
Marc Andreessen promotes an economic vision where crashing human wages through AI leads to greater productivity and lower consumer prices. He argues this pain is necessary for a future of abundance, dismissing concerns of creating a permanent underclass as a mere side effect of progress.
Sentiment: Critical | Time to Impact: Immediate | Future Signal: AI will accelerate
AI Healthcare
[EDITORS NOTE: The prediction below is not from longevity experts. This study is: https://www.nature.com/articles/s43587-024-00702-3)
Anthropic CEO Predicts AI Advances Could Double Human Lifespans in Just Years
Source: pymnts.com | Link
Anthropic's CEO Dario Amodei asserts that rapid AI advancements could potentially double human lifespans within five to ten years. He emphasizes AI's transformative role in biology and workplace efficiency, while also acknowledging regulatory and practical challenges that may impede such progress.
Sentiment: Positive | Time to Impact: Short-term | Future Signal: AI will accelerate
AI and Robotics
Humanoid Robots: Discover The 16 Leading Manufacturers Driving Innovation
Source: forbes.com | Link
Sixteen major manufacturers are progressing in humanoid robots, projected to revolutionize various sectors, from healthcare to entertainment, with vast economic implications. Analysts anticipate the market could reach $38 billion by 2035, highlighting the transformative potential of affordable humanoid labor across global economies.
Sentiment: Positive | Time to Impact: Immediate | Future Signal: AI will accelerate
What’s next for robots
Source: technologyreview.com | Link
Public perceptions of robotics vary widely, ranging from enthusiasm to skepticism. Developments in humanoid robots and AI-driven learning signal potential growth in automation, despite challenges related to employment and operational integration. The future of robotics appears promising but faces significant practical hurdles.
Sentiment: Neutral | Time to Impact: Immediate | Future Signal: AI will evolve gradually
Meta's Yann LeCun Predicts New AI Architectures Paradigm and Decade of Robotics
Source: techcrunch.com | Link
Yann LeCun, Meta's chief AI scientist, forecasts a transformative AI architecture shift within three to five years, moving beyond generative AI. He also envisions the coming decade as pivotal for robotics, which will integrate AI advancements to enhance real-world applications.
Sentiment: Positive | Time to Impact: Immediate | Future Signal: AI will evolve gradually
AI in the Data Center
What Happens When We Can’t Just Build Bigger AI Datacenters Anymore?
Source: theregister.com | Link
As power constraints limit the expansion of AI datacenters, industry leaders advocate for a strategy of connecting existing facilities to create larger, virtual supercomputers. This shift may enhance efficiency but introduces challenges in latency and bandwidth management, necessitating innovative networking solutions.
Sentiment: Neutral | Time to Impact: Immediate | Future Signal: AI will evolve gradually
AI Carbon Footprint
Startup plugs AI datacenters into biogas-powered energy
Source: theregister.com | Link
Chainergy, a UK startup, leverages biogas to power AI datacenters, significantly reducing electricity costs by 44% while lowering emissions by 70%. This approach circumvents lengthy grid connection delays, addressing urgent energy supply issues, especially in the UK datacenter market.
Sentiment: Positive | Time to Impact: Short-term | Future Signal: AI will accelerate
Big Tech wants to plug data centers right into power plants. Utilities say it’s not fair
Source: apnews.com | Link
Tech giants are negotiating direct power plant access to meet escalating data center energy demands, primarily bypassing existing grid challenges. This raises concerns about potential power shortages for other users and equitable contributions to infrastructure costs, prompting regulatory scrutiny from the Federal Energy Regulatory Commission (FERC).
Sentiment: Critical | Time to Impact: Short-term | Future Signal: AI will accelerate
Interesting Papers & Articles on Applied AI
Evaluating Large Language Models: New Benchmarks for Comprehensive Assessment
Source: thoughtworks.com | Link
The article discusses the complexities of evaluating large language models (LLMs) which require new metrics beyond traditional metrics like precision and recall. It emphasizes the need for comprehensive evaluations that consider system interactions, decision-making processes, and real-world performance to build trust in LLM deployments.
Sentiment: Neutral | Time to Impact: Immediate | Future Signal: AI will evolve gradually
Sentiment Distribution
Positive: 8 (33.3%) | Negative: 4 (16.7%) | Critical: 9 (37.5%) | Neutral: 3 (12.5%)
Future Signals
AI is hype: 4 (16.7%)| AI will accelerate: 9 (37.5%) | AI will evolve gradually: 8 (33.3%) | AI is a bubble: 3 (12.5%)
Want to discuss AI, quantum, and other emerging technologies?
Join our Slack
About the Curious AI Newsletter
AI is hype. AI is a utopia. AI is a dystopia.
These are the narratives currently being told about AI. There are mixed signals for each scenario. The truth will lie somewhere in between. This newsletter provides a curated overview of positive and negative data points to support decision-makers in forecasts and horizon scanning. The selection of news items is intended to provide a cross-section of articles from across the spectrum of AI optimists, AI realists, and AI pessimists and showcase the impact of AI across different domains and fields.
The news is curated by Oliver Rochford, Technologist, and former Gartner Research Director. AI (ChatGPT) is used in analysis and for summaries.