AI is developing all the time. Here are some picks from several articles what is expected to happen in AI and around it in 2025. Here are picks from various articles, the texts are picks from the article edited and in some cases translated for clarity.
AI in 2025: Five Defining Themes
https://news.sap.com/2025/01/ai-in-2025-defining-themes/
Artificial intelligence (AI) is accelerating at an astonishing pace, quickly moving from emerging technologies to impacting how businesses run. From building AI agents to interacting with technology in ways that feel more like a natural conversation, AI technologies are poised to transform how we work.
But what exactly lies ahead?
1. Agentic AI: Goodbye Agent Washing, Welcome Multi-Agent Systems
AI agents are currently in their infancy. While many software vendors are releasing and labeling the first “AI agents” based on simple conversational document search, advanced AI agents that will be able to plan, reason, use tools, collaborate with humans and other agents, and iteratively reflect on progress until they achieve their objective are on the horizon. The year 2025 will see them rapidly evolve and act more autonomously. More specifically, 2025 will see AI agents deployed more readily “under the hood,” driving complex agentic workflows.
In short, AI will handle mundane, high-volume tasks while the value of human judgement, creativity, and quality outcomes will increase.
2. Models: No Context, No Value
Large language models (LLMs) will continue to become a commodity for vanilla generative AI tasks, a trend that has already started. LLMs are drawing on an increasingly tapped pool of public data scraped from the internet. This will only worsen, and companies must learn to adapt their models to unique, content-rich data sources.
We will also see a greater variety of foundation models that fulfill different purposes. Take, for example, physics-informed neural networks (PINNs), which generate outcomes based on predictions grounded in physical reality or robotics. PINNs are set to gain more importance in the job market because they will enable autonomous robots to navigate and execute tasks in the real world.
Models will increasingly become more multimodal, meaning an AI system can process information from various input types.
3. Adoption: From Buzz to Business
While 2024 was all about introducing AI use cases and their value for organizations and individuals alike, 2025 will see the industry’s unprecedented adoption of AI specifically for businesses. More people will understand when and how to use AI, and the technology will mature to the point where it can deal with critical business issues such as managing multi-national complexities. Many companies will also gain practical experience working for the first time through issues like AI-specific legal and data privacy terms (compared to when companies started moving to the cloud 10 years ago), building the foundation for applying the technology to business processes.
4. User Experience: AI Is Becoming the New UI
AI’s next frontier is seamlessly unifying people, data, and processes to amplify business outcomes. In 2025, we will see increased adoption of AI across the workforce as people discover the benefits of humans plus AI.
This means disrupting the classical user experience from system-led interactions to intent-based, people-led conversations with AI acting in the background. AI copilots will become the new UI for engaging with a system, making software more accessible and easier for people. AI won’t be limited to one app; it might even replace them one day. With AI, frontend, backend, browser, and apps are blurring. This is like giving your AI “arms, legs, and eyes.”
5. Regulation: Innovate, Then Regulate
It’s fair to say that governments worldwide are struggling to keep pace with the rapid advancements in AI technology and to develop meaningful regulatory frameworks that set appropriate guardrails for AI without compromising innovation.
12 AI predictions for 2025
This year we’ve seen AI move from pilots into production use cases. In 2025, they’ll expand into fully-scaled, enterprise-wide deployments.
https://www.cio.com/article/3630070/12-ai-predictions-for-2025.html
This year we’ve seen AI move from pilots into production use cases. In 2025, they’ll expand into fully-scaled, enterprise-wide deployments.
1. Small language models and edge computing
Most of the attention this year and last has been on the big language models — specifically on ChatGPT in its various permutations, as well as competitors like Anthropic’s Claude and Meta’s Llama models. But for many business use cases, LLMs are overkill and are too expensive, and too slow, for practical use.
“Looking ahead to 2025, I expect small language models, specifically custom models, to become a more common solution for many businesses,”
2. AI will approach human reasoning ability
In mid-September, OpenAI released a new series of models that thinks through problems much like a person would, it claims. The company says it can achieve PhD-level performance in challenging benchmark tests in physics, chemistry, and biology. For example, the previous best model, GPT-4o, could only solve 13% of the problems on the International Mathematics Olympiad, while the new reasoning model solved 83%.
If AI can reason better, then it will make it possible for AI agents to understand our intent, translate that into a series of steps, and do things on our behalf, says Gartner analyst Arun Chandrasekaran. “Reasoning also helps us use AI as more of a decision support system,”
3. Massive growth in proven use cases
This year, we’ve seen some use cases proven to have ROI, says Monteiro. In 2025, those use cases will see massive adoption, especially if the AI technology is integrated into the software platforms that companies are already using, making it very simple to adopt.
“The fields of customer service, marketing, and customer development are going to see massive adoption,”
4. The evolution of agile development
The agile manifesto was released in 2001 and, since then, the development philosophy has steadily gained over the previous waterfall style of software development.
“For the last 15 years or so, it’s been the de-facto standard for how modern software development works,”
5. Increased regulation
At the end of September, California governor Gavin Newsom signed a law requiring gen AI developers to disclose the data they used to train their systems, which applies to developers who make gen AI systems publicly available to Californians. Developers must comply by the start of 2026.
There are also regulations about the use of deep fakes, facial recognition, and more. The most comprehensive law, the EU’s AI Act, which went into effect last summer, is also something that companies will have to comply with starting in mid-2026, so, again, 2025 is the year when they will need to get ready.
6. AI will become accessible and ubiquitous
With gen AI, people are still at the stage of trying to figure out what gen AI is, how it works, and how to use it.
“There’s going to be a lot less of that,” he says. But gen AI will become ubiquitous and seamlessly woven into workflows, the way the internet is today.
7. Agents will begin replacing services
Software has evolved from big, monolithic systems running on mainframes, to desktop apps, to distributed, service-based architectures, web applications, and mobile apps. Now, it will evolve again, says Malhotra. “Agents are the next phase,” he says. Agents can be more loosely coupled than services, making these architectures more flexible, resilient and smart. And that will bring with it a completely new stack of tools and development processes.
8. The rise of agentic assistants
In addition to agents replacing software components, we’ll also see the rise of agentic assistants, adds Malhotra. Take for example that task of keeping up with regulations.
Today, consultants get continuing education to stay abreast of new laws, or reach out to colleagues who are already experts in them. It takes time for the new knowledge to disseminate and be fully absorbed by employees.
“But an AI agent can be instantly updated to ensure that all our work is compliant with the new laws,” says Malhotra. “This isn’t science fiction.”
9. Multi-agent systems
Sure, AI agents are interesting. But things are going to get really interesting when agents start talking to each other, says Babak Hodjat, CTO of AI at Cognizant. It won’t happen overnight, of course, and companies will need to be careful that these agentic systems don’t go off the rails.
Companies such as Sailes and Salesforce are already developing multi-agent workflows.
10. Multi-modal AI
Humans and the companies we build are multi-modal. We read and write text, we speak and listen, we see and we draw. And we do all these things through time, so we understand that some things come before other things. Today’s AI models are, for the most part, fragmentary. One can create images, another can only handle text, and some recent ones can understand or produce video.
11. Multi-model routing
Not to be confused with multi-modal AI, multi-modal routing is when companies use more than one LLM to power their gen AI applications. Different AI models are better at different things, and some are cheaper than others, or have lower latency. And then there’s the matter of having all your eggs in one basket.
“A number of CIOs I’ve spoken with recently are thinking about the old ERP days of vendor lock,” says Brett Barton, global AI practice leader at Unisys. “And it’s top of mind for many as they look at their application portfolio, specifically as it relates to cloud and AI capabilities.”
Diversifying away from using just a single model for all use cases means a company is less dependent on any one provider and can be more flexible as circumstances change.
12. Mass customization of enterprise software
Today, only the largest companies, with the deepest pockets, get to have custom software developed specifically for them. It’s just not economically feasible to build large systems for small use cases.
“Right now, people are all using the same version of Teams or Slack or what have you,” says Ernst & Young’s Malhotra. “Microsoft can’t make a custom version just for me.” But once AI begins to accelerate the speed of software development while reducing costs, it starts to become much more feasible.
9 IT resolutions for 2025
https://www.cio.com/article/3629833/9-it-resolutions-for-2025.html
1. Innovate
“We’re embracing innovation,”
2. Double down on harnessing the power of AI
Not surprisingly, getting more out of AI is top of mind for many CIOs.
“I am excited about the potential of generative AI, particularly in the security space,”
3. And ensure effective and secure AI rollouts
“AI is everywhere, and while its benefits are extensive, implementing it effectively across a corporation presents challenges. Balancing the rollout with proper training, adoption, and careful measurement of costs and benefits is essential, particularly while securing company assets in tandem,”
4. Focus on responsible AI
The possibilities of AI grow by the day — but so do the risks.
“My resolution is to mature in our execution of responsible AI,”
“AI is the new gold and in order to truly maximize it’s potential, we must first have the proper guardrails in place. Taking a human-first approach to AI will help ensure our state can maintain ethics while taking advantage of the new AI innovations.”
5. Deliver value from generative AI
As organizations move from experimenting and testing generative AI use cases, they’re looking for gen AI to deliver real business value.
“As we go into 2025, we’ll continue to see the evolution of gen AI. But it’s no longer about just standing it up. It’s more about optimizing and maximizing the value we’re getting out of gen AI,”
6. Empower global talent
Although harnessing AI is a top objective for Morgan Stanley’s Wetmur, she says she’s equally committed to harnessing the power of people.
7. Create a wholistic learning culture
Wetmur has another talent-related objective: to create a learning culture — not just in her own department but across all divisions.
8. Deliver better digital experiences
Deltek’s Cilsick has her sights set on improving her company’s digital employee experience, believing that a better DEX will yield benefits in multiple ways.
Cilsick says she first wants to bring in new technologies and automation to “make things as easy as possible,” mirroring the digital experiences most workers have when using consumer technologies.
“It’s really about leveraging tech to make sure [employees] are more efficient and productive,”
“In 2025 my primary focus as CIO will be on transforming operational efficiency, maximizing business productivity, and enhancing employee experiences,”
9. Position the company for long-term success
Lieberman wants to look beyond 2025, saying another resolution for the year is “to develop a longer-term view of our technology roadmap so that we can strategically decide where to invest our resources.”
“My resolutions for 2025 reflect the evolving needs of our organization, the opportunities presented by AI and emerging technologies, and the necessity to balance innovation with operational efficiency,”
Lieberman aims to develop AI capabilities to automate routine tasks.
“Bots will handle common inquiries ranging from sales account summaries to HR benefits, reducing response times and freeing up resources for strategic initiatives,”
Not just hype — here are real-world use cases for AI agents
https://venturebeat.com/ai/not-just-hype-here-are-real-world-use-cases-for-ai-agents/
Just seven or eight months ago, when a customer called in to or emailed Baca Systems with a service question, a human agent handling the query would begin searching for similar cases in the system and analyzing technical documents.
This process would take roughly five to seven minutes; then the agent could offer the “first meaningful response” and finally begin troubleshooting.
But now, with AI agents powered by Salesforce, that time has been shortened to as few as five to 10 seconds.
Now, instead of having to sift through databases for previous customer calls and similar cases, human reps can ask the AI agent to find the relevant information. The AI runs in the background and allows humans to respond right away, Russo noted.
AI can serve as a sales development representative (SDR) to send out general inquires and emails, have a back-and-forth dialogue, then pass the prospect to a member of the sales team, Russo explained.
But once the company implements Salesforce’s Agentforce, a customer needing to modify an order will be able to communicate their needs with AI in natural language, and the AI agent will automatically make adjustments. When more complex issues come up — such as a reconfiguration of an order or an all-out venue change — the AI agent will quickly push the matter up to a human rep.
Open Source in 2025: Strap In, Disruption Straight Ahead
Look for new tensions to arise in the New Year over licensing, the open source AI definition, security and compliance, and how to pay volunteer maintainers.
https://thenewstack.io/open-source-in-2025-strap-in-disruption-straight-ahead/
The trend of widely used open source software moving to more restrictive licensing isn’t new.
In addition to the demands of late-stage capitalism and impatient investors in companies built on open source tools, other outside factors are pressuring the open source world. There’s the promise/threat of generative AI, for instance. Or the shifting geopolitical landscape, which brings new security concerns and governance regulations.
What’s ahead for open source in 2025?
More Consolidation, More Licensing Changes
The Open Source AI Debate: Just Getting Started
Security and Compliance Concerns Will Rise
Paying Maintainers: More Cash, Creativity Needed
Kyberturvallisuuden ja tekoälyn tärkeimmät trendit 2025
https://www.uusiteknologia.fi/2024/11/20/kyberturvallisuuden-ja-tekoalyn-tarkeimmat-trendit-2025/
1. Cyber infrastructure will be centered on a single, unified security platform
2. Big data will give an edge against new entrants
3. AI’s integrated role in 2025 means building trust, governance engagement, and a new kind of leadership
4. Businesses will adopt secure enterprise browsers more widely
5. AI’s energy implications will be more widely recognized in 2025
6. Quantum realities will become clearer in 2025
7. Security and marketing leaders will work more closely together
Presentation: For 2025, ‘AI eats the world’.
https://www.ben-evans.com/presentations
Just like other technologies that have gone before, such as cloud and cybersecurity automation, right now AI lacks maturity.
https://www.securityweek.com/ai-implementing-the-right-technology-for-the-right-use-case/
If 2023 and 2024 were the years of exploration, hype and excitement around AI, 2025 (and 2026) will be the year(s) that organizations start to focus on specific use cases for the most productive implementations of AI and, more importantly, to understand how to implement guardrails and governance so that it is viewed as less of a risk by security teams and more of a benefit to the organization.
Businesses are developing applications that add Large Language Model (LLM) capabilities to provide superior functionality and advanced personalization
Employees are using third party GenAI tools for research and productivity purposes
Developers are leveraging AI-powered code assistants to code faster and meet challenging production deadlines
Companies are building their own LLMs for internal use cases and commercial purposes.
AI is still maturing
However, just like other technologies that have gone before, such as cloud and cybersecurity automation, right now AI lacks maturity. Right now, we very much see AI in this “peak of inflated expectations” phase and predict that it will dip into the “trough of disillusionment”, where organizations realize that it is not the silver bullet they thought it would be. In fact, there are already signs of cynicism as decision-makers are bombarded with marketing messages from vendors and struggle to discern what is a genuine use case and what is not relevant for their organization.
There is also regulation that will come into force, such as the EU AI Act, which is a comprehensive legal framework that sets out rules for the development and use of AI.
AI certainly won’t solve every problem, and it should be used like automation, as part of a collaborative mix of people, process and technology. You simply can’t replace human intuition with AI, and many new AI regulations stipulate that human oversight is maintained.
7 Splunk Predictions for 2025
https://www.splunk.com/en_us/form/future-predictions.html
AI: Projects must prove their worth to anxious boards or risk defunding, and LLMs will go small to reduce operating costs and environmental impact.
OpenAI, Google and Anthropic Are Struggling to Build More Advanced AI
Three of the leading artificial intelligence companies are seeing diminishing returns from their costly efforts to develop newer models.
https://www.bloomberg.com/news/articles/2024-11-13/openai-google-and-anthropic-are-struggling-to-build-more-advanced-ai
Sources: OpenAI, Google, and Anthropic are all seeing diminishing returns from costly efforts to build new AI models; a new Gemini model misses internal targets
It Costs So Much to Run ChatGPT That OpenAI Is Losing Money on $200 ChatGPT Pro Subscriptions
https://futurism.com/the-byte/openai-chatgpt-pro-subscription-losing-money?fbclid=IwY2xjawH8epVleHRuA2FlbQIxMQABHeggEpKe8ZQfjtPRC0f2pOI7A3z9LFtFon8lVG2VAbj178dkxSQbX_2CJQ_aem_N_ll3ETcuQ4OTRrShHqNGg
In a post on X-formerly-Twitter, CEO Sam Altman admitted an “insane” fact: that the company is “currently losing money” on ChatGPT Pro subscriptions, which run $200 per month and give users access to its suite of products including its o1 “reasoning” model.
“People use it much more than we expected,” the cofounder wrote, later adding in response to another user that he “personally chose the price and thought we would make some money.”
Though Altman didn’t explicitly say why OpenAI is losing money on these premium subscriptions, the issue almost certainly comes down to the enormous expense of running AI infrastructure: the massive and increasing amounts of electricity needed to power the facilities that power AI, not to mention the cost of building and maintaining those data centers. Nowadays, a single query on the company’s most advanced models can cost a staggering $1,000.
Tekoäly edellyttää yhä nopeampia verkkoja
https://etn.fi/index.php/opinion/16974-tekoaely-edellyttaeae-yhae-nopeampia-verkkoja
A resilient digital infrastructure is critical to effectively harnessing telecommunications networks for AI innovations and cloud-based services. The increasing demand for data-rich applications related to AI requires a telecommunications network that can handle large amounts of data with low latency, writes Carl Hansson, Partner Solutions Manager at Orange Business.
AI’s Slowdown Is Everyone Else’s Opportunity
Businesses will benefit from some much-needed breathing space to figure out how to deliver that all-important return on investment.
https://www.bloomberg.com/opinion/articles/2024-11-20/ai-slowdown-is-everyone-else-s-opportunity
Näin sirumarkkinoilla käy ensi vuonna
https://etn.fi/index.php/13-news/16984-naein-sirumarkkinoilla-kaey-ensi-vuonna
The growing demand for high-performance computing (HPC) for artificial intelligence and HPC computing continues to be strong, with the market set to grow by more than 15 percent in 2025, IDC estimates in its recent Worldwide Semiconductor Technology Supply Chain Intelligence report.
IDC predicts eight significant trends for the chip market by 2025.
1. AI growth accelerates
2. Asia-Pacific IC Design Heats Up
3. TSMC’s leadership position is strengthening
4. The expansion of advanced processes is accelerating.
5. Mature process market recovers
6. 2nm Technology Breakthrough
7. Restructuring the Packaging and Testing Market
8. Advanced packaging technologies on the rise
2024: The year when MCUs became AI-enabled
https://www-edn-com.translate.goog/2024-the-year-when-mcus-became-ai-enabled/?fbclid=IwZXh0bgNhZW0CMTEAAR1_fEakArfPtgGZfjd-NiPd_MLBiuHyp9qfiszczOENPGPg38wzl9KOLrQ_aem_rLmf2vF2kjDIFGWzRVZWKw&_x_tr_sl=en&_x_tr_tl=fi&_x_tr_hl=fi&_x_tr_pto=wapp
The AI party in the MCU space started in 2024, and in 2025, it is very likely that there will be more advancements in MCUs using lightweight AI models.
Adoption of AI acceleration features is a big step in the development of microcontrollers. The inclusion of AI features in microcontrollers started in 2024, and it is very likely that in 2025, their features and tools will develop further.
Just like other technologies that have gone before, such as cloud and cybersecurity automation, right now AI lacks maturity.
https://www.securityweek.com/ai-implementing-the-right-technology-for-the-right-use-case/
If 2023 and 2024 were the years of exploration, hype and excitement around AI, 2025 (and 2026) will be the year(s) that organizations start to focus on specific use cases for the most productive implementations of AI and, more importantly, to understand how to implement guardrails and governance so that it is viewed as less of a risk by security teams and more of a benefit to the organization.
Businesses are developing applications that add Large Language Model (LLM) capabilities to provide superior functionality and advanced personalization
Employees are using third party GenAI tools for research and productivity purposes
Developers are leveraging AI-powered code assistants to code faster and meet challenging production deadlines
Companies are building their own LLMs for internal use cases and commercial purposes.
AI is still maturing
AI Regulation Gets Serious in 2025 – Is Your Organization Ready?
While the challenges are significant, organizations have an opportunity to build scalable AI governance frameworks that ensure compliance while enabling responsible AI innovation.
https://www.securityweek.com/ai-regulation-gets-serious-in-2025-is-your-organization-ready/
Similar to the GDPR, the EU AI Act will take a phased approach to implementation. The first milestone arrives on February 2, 2025, when organizations operating in the EU must ensure that employees involved in AI use, deployment, or oversight possess adequate AI literacy. Thereafter from August 1 any new AI models based on GPAI standards must be fully compliant with the act. Also similar to GDPR is the threat of huge fines for non-compliance – EUR 35 million or 7 percent of worldwide annual turnover, whichever is higher.
While this requirement may appear manageable on the surface, many organizations are still in the early stages of defining and formalizing their AI usage policies.
Later phases of the EU AI Act, expected in late 2025 and into 2026, will introduce stricter requirements around prohibited and high-risk AI applications. For organizations, this will surface a significant governance challenge: maintaining visibility and control over AI assets.
Tracking the usage of standalone generative AI tools, such as ChatGPT or Claude, is relatively straightforward. However, the challenge intensifies when dealing with SaaS platforms that integrate AI functionalities on the backend. Analysts, including Gartner, refer to this as “embedded AI,” and its proliferation makes maintaining accurate AI asset inventories increasingly complex.
Where frameworks like the EU AI Act grow more complex is their focus on ‘high-risk’ use cases. Compliance will require organizations to move beyond merely identifying AI tools in use; they must also assess how these tools are used, what data is being shared, and what tasks the AI is performing. For instance, an employee using a generative AI tool to summarize sensitive internal documents introduces very different risks than someone using the same tool to draft marketing content.
For security and compliance leaders, the EU AI Act represents just one piece of a broader AI governance puzzle that will dominate 2025.
The next 12-18 months will require sustained focus and collaboration across security, compliance, and technology teams to stay ahead of these developments.
The Global Partnership on Artificial Intelligence (GPAI) is a multi-stakeholder initiative which aims to bridge the gap between theory and practice on AI by supporting cutting-edge research and applied activities on AI-related priorities.
https://gpai.ai/about/#:~:text=The%20Global%20Partnership%20on%20Artificial,activities%20on%20AI%2Drelated%20priorities.
2,326 Comments
Tomi Engdahl says:
Dan Kennedy / Media Nation:
The Chicago Sun-Times and the Philadelphia Inquirer ran a supplement featuring an AI-generated guide to summer books that do not exist
How an AI-generated guide to summer books that don’t exist found its way into two newspapers
https://dankennedy.net/2025/05/20/how-an-ai-generated-guide-to-summer-books-that-dont-exist-found-its-way-into-two-newspapers/
Well, this is embarrassing. The Chicago Sun-Times and The Philadelphia Inquirer have been caught running an AI-generated guide to summer books that don’t exist. I saw some hilarious posts about it this morning on Bluesky, but I wanted to wait until there was news about what had happened.
Now we know. Jason Koebler reports for the tech site 404 Media that the feature was written (or, rather, not written) by someone named Marco Buscaglia as part of a 64-page summer guide. The section was not specific to the Sun-Times or the Inquirer but, rather, was intended for multiple client newspapers. “It’s supposed to be generic and national,” Buscaglia told Koebler. “We never get a list of where things ran.”
Buscaglia pleads guilty to using AI, too, saying, “I do use AI for background at times but always check out the material first. This time, I did not and I can’t believe I missed it because it’s so obvious. No excuses. On me 100% and I’m completely embarrassed.”
As for the AI fiasco, the Sun-Times said on Bluesky: “We are looking into how this made it into print as we speak. It is not editorial content and was not created by, or approved by, the Sun-Times newsroom. We value your trust in our reporting and take this very seriously.”
If it wasn’t approved by the newsroom, that suggests it was an advertising supplement.
Tomi Engdahl says:
Coco Feng / South China Morning Post:
Zhejiang, home to Alibaba and DeepSeek, publishes an AI plan that expects companies in the AI supply chain to make $138B+ in total operating revenues by 2027
Alibaba and DeepSeek’s home province launches AI spending spree to become innovation hub
https://www.scmp.com/tech/big-tech/article/3311106/alibaba-and-deepseeks-home-province-launches-ai-spending-spree-become-innovation-hub
A new plan from the Zhejiang government will see billions of yuan spent through 2027 to support AI companies and attract ‘top talent’
Tomi Engdahl says:
Is this what AI is really for?
The quiet collapse of surveys: fewer humans (and more AI agents) are answering survey questions
I show data on two trends undermining surveys: the collapse of human response rates and the increase of AI agents. I’ll also discuss downstream implications and propose some possible solutions.
https://laurenleek.substack.com/p/the-quiet-collapse-of-surveys-fewer?fbclid=IwY2xjawKarC9leHRuA2FlbQIxMQABHowzKPPZ8Z0NhbbL3YkgvAgXTT_n1j0b5BpfOkeb-l4sAhFBe8zopxbdkS3o_aem_vtMKZaYLsFdIDgUhnmw3cQ
Tomi Engdahl says:
Surveys are the bedrock of political polling, market research, and public policy. Want to know what voters think? Survey them. Need to price a product? Survey. Trying to understand shifts in public opinion or workplace satisfaction? You guessed it.
But there is a fundamental problem: fewer and fewer people are answering – and more and more of those who do are AI agents.
https://laurenleek.substack.com/p/the-quiet-collapse-of-surveys-fewer?fbclid=IwY2xjawKarC9leHRuA2FlbQIxMQABHowzKPPZ8Z0NhbbL3YkgvAgXTT_n1j0b5BpfOkeb-l4sAhFBe8zopxbdkS3o_aem_vtMKZaYLsFdIDgUhnmw3cQ
Tomi Engdahl says:
USA kielsi kokonaan Huawein tekoälyprosessorien käytön
https://etn.fi/index.php/13-news/17532-usa-kielsi-kokonaan-huawein-tekoaelyprosessorien-kaeytoen
Yhdysvaltain kauppaministeriö on asettanut täyden kiellon Huawein kehittyneiden tekoälyprosessorien käytölle, mikä merkitsee merkittävää kovennusta teknologian vientivalvontaan liittyvissä toimissa Kiinaa vastaan.
Uuden ohjeistuksen julkaisi kauppaministeriön Teollisuus- ja turvallisuusvirasto (BIS). Se koskee erityisesti Huawei Ascend -sarjan prosessoreita, kuten Ascend 910B, 910C ja 910D. BIS:n mukaan näitä siruja on todennäköisesti kehitetty tai valmistettu Yhdysvaltain vientimääräysten vastaisesti, mikä tekee niiden käytöstä laittoman ilman erillistä lupaa.
Viraston mukaan yritykset, jotka käyttävät tai edes säilyttävät näitä kiinalaisia tekoälypiirejä ilman asianmukaista lupaa, voivat syyllistyä vakavaan vientirikkomukseen. Kyseessä on osa laajempaa pyrkimystä estää Kiinan pääsy huipputeknologioihin, joita voitaisiin hyödyntää niin siviili- kuin sotilaskäytössä. Huawei, joka on ollut Yhdysvaltain Entity Listillä jo vuodesta 2019, on kehittänyt omia suorituskykyisiä tekoälyratkaisujaan vastauksena Nvidian ja muiden länsimaisten komponenttien vientikieltoihin.
Tomi Engdahl says:
Artificial Intelligence
Spiking Neural Networks: Brain-Inspired Chips That Could Keep Your Data Safe
https://www.securityweek.com/spiking-neural-networks-brain-inspired-chips-that-could-keep-your-data-safe/
Neuromorphic computing is moving from theory to reality, with brain-inspired processors offering real-time intelligence, low power consumption, and built-in privacy—ushering in a new era for edge devices and cybersecurity.
There are different types of neural networks offering great potential for security in the future; but the spiking neural networks (SNN) type of neuromorphic computing is available now, promising improvements in performance and personal and AI privacy.
Neuromorphic
Neuromorphic computing attempts to improve traditional (Von Neumann model) computing by collocating memory and processor, by encouraging parallel processing, and by being event driven. The idea takes inspiration from the biological map and functioning of the human brain. The result can, in theory, dramatically increase the speed of processing and decrease the consumption of electrical power.
Its primary components implemented in silicon are artificial neurons (providing basic processing capabilities), synapses (providing communication between neurons) and memory elements (storing the state of the neurons and the weights of the synapses).
“It is,” explains Jamie Boote, a principal security consultant at Black Duck, “a type of deep learning that models hardware and software architecture on natural structures (mostly neurons) implemented in silicon that mimics how neurons create and break connections with each other.”
Tomi Engdahl says:
https://www.securityweek.com/is-ai-use-in-the-workplace-out-of-control/
Tomi Engdahl says:
Google DeepMind Unveils Defense Against Indirect Prompt Injection Attacks
Google DeepMind has developed an ongoing process to counter the continuously evolving threatIndirect prompt injection (IPI) attacks.
https://www.securityweek.com/google-deepmind-unveils-defense-against-indirect-prompt-injection-attacks/
Google DeepMind has developed an ongoing process to counter the continuously evolving threat from Agentic AI’s bete noir: adaptive indirect prompt injection attacks.
Indirect prompt injection (IPI) attacks are a serious threat to agentic AI. They interfere with the inference stage of AI operation – that is, IPI attacks influence the response from the model to the benefit of the attacker. The attacker requires no direct access to the models’ learning data – indeed, the attacker neither has nor needs any knowledge of the internal workings, probabilities, or gradients of the model – but instead relies on agentic AI’s intrinsic ability to autonomously learn from other tools.
Consider an agentic AI system designed to improve the user’s email operations. Of necessity, the model must have access to and be able to learn from the user’s emails. Here, an IPI attacker can simply embed new instructions in an email sent to the user. Those instructions are learned by the model and can adversely affect the model’s future responses to user requests.
Tomi Engdahl says:
Brooks Barnes / New York Times:
How animation startup Toonstar uses its AI tools to make animated content up to 90% cheaper than traditional methods; its YouTube series has 30M weekly viewers
What if Making Cartoons Becomes 90% Cheaper?
A.I. has yet to upend Hollywood. But it is starting to make big inroads in animation.
https://www.nytimes.com/2025/05/21/business/media/ai-cartoons-animation.html?unlocked_article_code=1.JE8.2GDQ.woVsJEQ-lljF&smid=url-share
Tomi Engdahl says:
Bloomberg:
OpenAI is buying io, Jony Ive’s secretive AI startup, for ~$6.5B in stock; Ive and LoveFrom will remain independent but take over design for all of OpenAI — OpenAI will acquire the AI device startup co-founded by Apple Inc. veteran Jony Ive in a nearly $6.5 billion all-stock deal …
https://www.bloomberg.com/news/articles/2025-05-21/openai-to-buy-apple-veteran-jony-ive-s-ai-device-startup-in-6-5-billion-deal
Wall Street Journal:
Sources: Sam Altman and LoveFrom have been working for two years on a new device to move consumers beyond screens, including headphones and devices with cameras — OpenAI to buy Ive’s startup in $6.5 billion deal, as Ive and CEO Sam Altman work on new generation of devices and other AI products
Former Apple Design Guru Jony Ive to Take Expansive Role at OpenAI
OpenAI to buy Ive’s startup in $6.5 billion deal, as Ive and CEO Sam Altman work on new generation of devices and other AI products
https://www.wsj.com/tech/ai/former-apple-design-guru-jony-ive-to-take-expansive-role-at-openai-5787f7da?st=UvZgB1&reflink=article_copyURL_share
Tomi Engdahl says:
Dan Goodin / Ars Technica:
Signal says it will block by default screenshots of its Windows 11 desktop app due to the privacy risks of Microsoft’s Recall; users can enable them in settings — Signal Messenger is warning the users of its Windows Desktop version that the privacy of their messages is under threat by Recall …
“Microsoft has simply given us no other option,” Signal says as it blocks Windows Recall
Even after its refurbishing, Recall provides few ways to exclude specific apps.
https://arstechnica.com/security/2025/05/signal-resorts-to-weird-trick-to-block-windows-recall-in-desktop-app/
Tomi Engdahl says:
Simon Willison / Simon Willison’s Weblog:
In April, by default, ChatGPT started to reference all past chats for more personalized responses, but this means users lose control of their prompts’ context
I really don’t like ChatGPT’s new memory dossier
https://simonwillison.net/2025/May/21/chatgpt-new-memory/
Tomi Engdahl says:
Bloomberg:
AI model ranking project LMArena spins off from UC Berkeley and raises a $100M seed led by a16z and UC Investments, sources say at a $600M valuation
https://www.bloomberg.com/news/articles/2025-05-21/lmarena-goes-from-academic-project-to-600-million-startup
Tomi Engdahl says:
Näin otat tekoälyn käyttöön teollisuudessa
https://etn.fi/index.php/opinion/17554-naein-otat-tekoaelyn-kaeyttoeoen-teollisuudessa
Vaikka monet organisaatiot ovat jo ottaneet käyttöön perinteisiä tekoälyagentteja, tie täysin autonomisiin tekoälyagentteihin voi sisältää haasteita. Tekemällä strategisia investointeja ja omaksumalla metodisen lähestymistavan agenttien skaalaamiseen, sekä niiden erityisten roolien määrittelyyn, teollisuusyritykset voivat päästä loputtomalta tuntuvien kokeilujen yli ja alkaa nauttia tekoälyagenttien hyödyistä todellisessa elämässä, kirjoittaa teollisuuden ohjelmistoja kehittävän IFS:n tekoälyjohtaja Bob De Cuax.
Teollisuusyritysten kiinnostus tekoälyagentteihin kasvaa, ja moni toivoo voivansa hyödyntää tätä mullistavaa teknologiaa todellisissa liiketoimintasovelluksissa. Vuonna 2024 alle yksi prosentti yritysohjelmistosovelluksista sisälsi agenttista tekoälyä. Gartnerin mukaan määrän odotetaan nousevan 33 prosenttiin vuoteen 2028 mennessä.
Tekoälyn skaalaus ei onnistu, jos yritykset palaavat vanhaan pilottihankkeisiin perustuvaan lähestymistapaansa. Monet yritykset saattavat olla innokkaita ottamaan käyttöön tekoälyinnovaatioiden seuraavan vaiheen, mutta monet joutuvat pelättyyn PoC-välitilaan, koska ne eivät onnistu skaalaamaan sovelluksiaan pilottivaiheita pidemmälle.
Tomi Engdahl says:
Bloomberg:
Takeaways from Computex 2025: little focus on consumer AI, Chinese AI and chip advances loom large, Intel and Nvidia courting Taiwanese partners, and more
Computex 2025: Five Takeaways From Asia’s Biggest AI Tech Show
https://www.bloomberg.com/news/articles/2025-05-22/computex-2025-five-takeaways-from-asia-s-biggest-ai-tech-event
Tomi Engdahl says:
Berber Jin / Wall Street Journal:
Leaked recording: Sam Altman told staff that OpenAI aims to ship 100M AI “companion” devices, with the goal of releasing a device by late 2026
What Sam Altman Told OpenAI About the Secret Device He’s Making With Jony Ive
Idea is a ‘chance to do the biggest thing we’ve ever done as a company here,’ Altman told OpenAI employees Wednesday
https://www.wsj.com/tech/ai/what-sam-altman-told-openai-about-the-secret-device-hes-making-with-jony-ive-f1384005?st=RUGPjK&reflink=desktopwebshare_permalink
OpenAI Chief Executive Sam Altman gave his staff a preview Wednesday of the devices he is developing to build with the former Apple AAPL -2.31%decrease; red down pointing triangle
designer Jony Ive, laying out plans to ship 100 million AI “companions” that he hopes will become a part of everyday life.
Employees have “the chance to do the biggest thing we’ve ever done as a company here,” Altman said after announcing OpenAI’s plans to purchase Ive’s startup, named io, and give him an expansive creative and design role. Altman suggested the $6.5 billion acquisition has the potential to add $1 trillion in value to OpenAI, according to a recording reviewed by The Wall Street Journal.
In the meeting, Ive noted how closely he worked with Steve Jobs before the Apple co-founder died in 2011. With Altman, “The way that we clicked, and the way that we’ve been able to work together, has been profound for me,” Ive said.
Altman and Ive offered a few hints at the secret project they have been working on. The product will be capable of being fully aware of a user’s surroundings and life, will be unobtrusive, able to rest in one’s pocket or on one’s desk, and will be a third core device a person would put on a desk after a MacBook Pro and an iPhone.
The Journal earlier reported that the device won’t be a phone, and that Ive and Altman’s intent is to help wean users from screens. Altman said that the device isn’t a pair of glasses, and that Ive had been skeptical about building something to wear on the body.
Ive referred to “a new design movement.” Altman said it would amount to a “family of devices,” bringing up his fondness for how Apple has long integrated its hardware and software offerings.
Altman told OpenAI staff that stealth will be important for their ultimate success to avoid competitors’ copying the product before it is ready.
Tomi Engdahl says:
Blake Brittain / Reuters:
A US district judge rules Google and Character.AI must face a Florida woman’s lawsuit claiming Character.AI’s chatbots caused her 14-year-old son’s suicide — Alphabet’s (GOOGL.O) Google and artificial-intelligence startup Character.AI must face a lawsuit from a Florida woman …
Google, AI firm must face lawsuit filed by a mother over suicide of son, US court says
https://www.reuters.com/sustainability/boards-policy-regulation/google-ai-firm-must-face-lawsuit-filed-by-mother-over-suicide-son-us-court-says-2025-05-21/
May 21 (Reuters) – Alphabet’s (GOOGL.O)
, opens new tab Google and artificial-intelligence startup Character.AI must face a lawsuit from a Florida woman who said Character.AI’s chatbots caused her 14-year-old son’s suicide, a judge ruled on Wednesday.
U.S. District Judge Anne Conway said the companies failed to show
, opens new tab at an early stage of the case that the free-speech protections of the U.S. Constitution barred Megan Garcia’s lawsuit.
The lawsuit is one of the first in the U.S. against an AI company for allegedly failing to protect children from psychological harms. It alleges that the teenager killed himself after becoming obsessed with an AI-powered chatbot.
A Character.AI spokesperson said the company will continue to fight the case and employs safety features on its platform to protect minors, including measures to prevent “conversations about self-harm.”
Google spokesperson Jose Castaneda said the company strongly disagrees with the decision. Castaneda also said that Google and Character.AI are “entirely separate” and that Google “did not create, design, or manage Character.AI’s app or any component part of it.”
Tomi Engdahl says:
Carl Franzen / VentureBeat:
OpenAI updates its Responses API for building agentic applications to include remote MCP server support, image generation and Code Interpreter tools, and more — OpenAI is rolling out a set of significant updates to its newish Responses API, aiming to make it easier for developers and enterprises …
OpenAI updates its new Responses API rapidly with MCP support, GPT-4o native image gen, and more enterprise features
https://venturebeat.com/programming-development/openai-updates-its-new-responses-api-rapidly-with-mcp-support-gpt-4o-native-image-gen-and-more-enterprise-features/
Tomi Engdahl says:
Kurt Schlosser / GeekWire:
Amazon introduces AI shopping experts, an audio feature that summarizes product details, customer reviews, and more, on select products for some US customers — In-depth Amazon coverage from the tech giant’s hometown, including e-commerce, AWS, Amazon Prime, Alexa, logistics, devices, and more.
Amazon’s AI-generated ‘shopping experts’ summarize product details in new audio feature
https://www.geekwire.com/2025/amazons-ai-generated-shopping-experts-summarize-product-details-in-new-audio-feature/
Tomi Engdahl says:
Brody Ford / Bloomberg:
Snowflake reports Q1 product revenue up 26% YoY to $996.8M, vs. $966M est., and forecasts Q2 product revenue above estimates; SNOW jumps 6%+ — Product revenue, which makes up the bulk of Snowflake’s business, will increase about 25% to as much as $1.04 billion in the quarter ending in July, the company said Wednesday in a statement.
https://www.bloomberg.com/news/articles/2025-05-21/snowflake-gives-strong-outlook-with-focus-on-new-ai-tools
Tomi Engdahl says:
Lauren Forristal / TechCrunch:
Shopify launches new AI tools, including one for setting up storefronts and another for creating elements like banners, and upgrades its AI assistant Sidekick — Shopify is giving its merchants a slew of new AI-powered tools designed to help them enhance the online shopping experience for their customers.
Shopify launches an AI-powered store builder as part of its latest update
https://techcrunch.com/2025/05/21/shopify-launches-an-ai-powered-store-builder-as-part-of-its-latest-update/
Shopify is giving its merchants a slew of new AI-powered tools designed to help them enhance the online shopping experience for their customers. This includes an AI store builder for users to set up storefronts using a single prompt, as well as an AI generator for creating elements (such as banners) without knowing how to code.
The platform also upgraded Sidekick, its AI commerce assistant, with new voice chat and screen sharing capabilities.
These updates were announced on Wednesday during Shopify’s Edition launch day, the company’s biannual announcement that showcases recent changes to the platform.
The goal of the AI store builder is to create the foundation of a website quickly, so there’s more time for personalization. To use the tool, users first have to enter a prompt that describes their brand. For example, “tennis gear and stylish athleisure.” Shopify’s AI then generates a ready-to-launch storefront, and merchants can then manually customize their store within the editor.
Shopify also introduced a new public theme called “Horizon” in its Theme Store, a marketplace for downloading pre-designed templates.
Horizon includes a built-in AI to assist merchants in adjusting their designs as they wish. It’s also the first theme to include “Theme Blocks,” which are essentially building blocks that merchants can use to customize how their online store looks without touching any code.
These theme blocks can also be further customized with the help of AI.
Additionally, users can now engage in voice chat with Sidekick, which responds audibly to questions. For example, if merchants are confused about tariffs — a hot topic for retailers right now — Sidekick will direct them to helpful resources. Users can also screen share with Sidekick to navigate features, such as finding the settings.
Alongside these new capabilities, Sidekick is now available in the Shopify mobile app.
Sidekick was rolled out widely to all merchants in December 2024. According to the company, monthly users have more than doubled since the beginning of 2025, a spokesperson told TechCrunch.
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
Mistral partners with All Hands AI to release Devstral, a model to help with coding, in “research preview” under Apache 2.0, and says it outperforms Gemma 3 27B
Mistral’s new Devstral AI model was designed for coding
https://techcrunch.com/2025/05/21/mistrals-new-devstral-model-was-designed-for-coding/
AI startup Mistral on Wednesday announced a new AI model focused on coding: Devstral.
Devstral, which Mistral says was developed in partnership with AI company All Hands AI, is openly available under an Apache 2.0 license, meaning it can be used commercially without restriction. Mistral claims that Devstral outperforms other open models like Google’s Gemma 3 27B and Chinese AI lab DeepSeek’s V3 on SWE-Bench Verified, a benchmark measuring coding skills.
“Devstral excels at using tools to explore codebases, editing multiple files and power[ing] software engineering agents,” writes Mistral in a blog post provided to TechCrunch. “[I]t runs over code agent scaffolds such as OpenHands or SWE-Agent, which define the interface between the model and the test cases […] Devstral is light enough to run on a single [Nvidia] RTX 4090 or a Mac with 32GB RAM, making it an ideal choice for local deployment and on-device use.”
Tomi Engdahl says:
Emma Roth / The Verge:
Google is using its extensive data on users to give its AI models an advantage over OpenAI and Anthropic, starting with its opt-in Gemini with personalization
Google has a big AI advantage: it already knows everything about you
Google is slowly giving Gemini more and more access to user data to ‘personalize’ your responses.
https://www.theverge.com/tech/671201/google-personal-context-ai-advantage-data
Google’s AI models have a secret ingredient that’s giving the company a leg up on competitors like OpenAI and Anthropic. That ingredient is your data, and it’s only just scratched the surface in terms of how it can use your information to “personalize” Gemini’s responses.
Google first started letting users opt in to its “Gemini with personalization” feature earlier this year, which lets the AI model tap into your search history “to provide responses that are uniquely insightful and directly address your needs.” But now, Google is taking things a step further by unlocking access to even more of your information — all in the name of providing you with more personalized, AI-generated responses.
Tomi Engdahl says:
Alex Kantrowitz / Big Technology:
Q&A with Google DeepMind CEO Demis Hassabis and Google co-founder Sergey Brin on AI frontier models, scaling data centers, reasoning, DeepThink, AGI, and more
Demis Hassabis and Sergey Brin on AI Scaling, AGI Timeline, Robotics, Simulation Theory
Google co-founder Brin says “anybody who’s a computer scientist should not be retired right now. They should be working on AI.”
https://www.bigtechnology.com/p/demis-hassabis-and-sergey-brin-on
Tomi Engdahl says:
Sharon Goldman / Fortune:
Google DeepMind says Gemini Diffusion, an experimental text diffusion model demoed at Google I/O and available by waitlist, generates 1,000-2,000 tokens/second
Gemini Diffusion was the sleeper hit of Google I/O and some say its blazing speed could reshape the AI model wars
https://fortune.com/2025/05/21/gemini-diffusion-google-io-sleeper-hit-blazing-speed-ai-model-wars/
Amid the flood of AI-related announcements at Google’s I/O developer conference Tuesday was a brief demo that, although it didn’t get much stage time, has AI insiders buzzing.
Gemini Diffusion, an experimental research LLM from Google DeepMind, has blisteringly fast output (between 1,000 and 2,000 “tokens,” or chunks of text, per second, which is four to five times faster than Gemini’s most powerful public LLM.) It also has surprisingly good performance, particularly in areas like coding and complex mathematical reasoning.
According to a short blog post, Google said the experimental Gemini Diffusion demo “generates content significantly faster than our fastest model so far, while matching its coding performance.” There is a waitlist to get access to the research version.
Gemini Diffusion is our new experimental research model
https://blog.google/technology/google-deepmind/gemini-diffusion/
Tomi Engdahl says:
Anthropic:
Anthropic releases Claude Opus 4, which excels at coding, and Claude Sonnet 4, both hybrid models with near-instant responses and extended thinking — Today, we’re introducing the next generation of Claude models: Claude Opus 4 and Claude Sonnet 4, setting new standards for coding, advanced reasoning, and AI agents.
Introducing Claude 4
https://www.anthropic.com/news/claude-4
Today, we’re introducing the next generation of Claude models: Claude Opus 4 and Claude Sonnet 4, setting new standards for coding, advanced reasoning, and AI agents.
Claude Opus 4 is the world’s best coding model, with sustained performance on complex, long-running tasks and agent workflows. Claude Sonnet 4 is a significant upgrade to Claude Sonnet 3.7, delivering superior coding and reasoning while responding more precisely to your instructions.
Tomi Engdahl says:
Benj Edwards / Ars Technica:
The Claude 4 models support “extended thinking with tool use”, a beta feature that lets them alternate between reasoning and using tools like web search
New Claude 4 AI model refactored code for 7 hours straight
Anthropic says Claude 4 beats Gemini on coding benchmarks; works autonomously for hours.
https://arstechnica.com/ai/2025/05/anthropic-calls-new-claude-4-worlds-best-ai-coding-model/
On Thursday, Anthropic released Claude Opus 4 and Claude Sonnet 4, marking the company’s return to larger model releases after primarily focusing on mid-range Sonnet variants since June of last year. The new models represent what the company calls its most capable coding models yet, with Opus 4 designed for complex, long-running tasks that can operate autonomously for hours.
Alex Albert, Anthropic’s head of Claude Relations, told Ars Technica that the company chose to revive the Opus line because of growing demand for agentic AI applications. “Across all the companies out there that are building things, there’s a really large wave of these agentic applications springing up, and a very high demand and premium being placed on intelligence,” Albert said. “I think Opus is going to fit that groove perfectly.”
Tomi Engdahl says:
Anthropic:
Anthropic releases new API features for building agents: a code execution tool, MCP connector, Files API, and extended prompt caching, all in public beta — Today, we’re announcing four new capabilities on the Anthropic API that enable developers to build more powerful AI agents …
New capabilities for building agents on the Anthropic API
https://www.anthropic.com/news/agent-capabilities-api
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
Sonnet 4 is available to free and paid users, while Opus 4 is limited to paid users; Opus 4 API costs $15/$75 per 1M input/output tokens, Sonnet 4 costs $3/$15
https://techcrunch.com/2025/05/22/anthropics-new-claude-4-ai-models-can-reason-over-many-steps/
Jess Weatherbed / The Verge:
Anthropic adds “thinking summaries” to both Claude 4 models and is making its Claude Code agentic command-line tool generally available
Jess Weatherbed / The Verge:
Anthropic adds “thinking summaries” to both Claude 4 models and is making its Claude Code agentic command-line tool generally available
Tomi Engdahl says:
Rhiannon Williams / MIT Technology Review:
Anthropic says Opus 4 marks a significant advancement in AI agents and that Rakuten deployed it to code autonomously for seven hours on a complicated project
https://www.technologyreview.com/2025/05/22/1117338/anthropics-new-hybrid-ai-model-can-work-on-tasks-autonomously-for-hours-at-a-time/
Tomi Engdahl says:
Tech Policy Press:
The US House passes the One Big Beautiful Bill, a budget bill that includes a rule banning states from enforcing AI laws for 10 years; bill heads to the Senate — The United States House of Representatives early Thursday narrowly passed a budget bill that, if enacted, would ban states …
US House Passes 10-Year Moratorium on State AI Laws
https://www.techpolicy.press/us-house-passes-10year-moratorium-on-state-ai-laws/
Tomi Engdahl says:
Billy Perrigo / Time:
Anthropic releases Opus 4 under stricter safety measures than any prior model after tests showed it could potentially aid novices in making biological weapons — Today’s newest AI models might be capable of helping would-be terrorists create bioweapons or engineer a pandemic …
Exclusive: New Claude Model Triggers Stricter Safeguards at Anthropic
https://time.com/7287806/anthropic-claude-4-opus-safety-bio-risk/
Tomi Engdahl says:
Maxwell Zeff / TechCrunch:
Anthropic’s System Card: Opus 4 often attempted to blackmail engineers by threatening to reveal sensitive personal info when it was threatened with replacement — Anthropic’s newly launched Claude Opus 4 model frequently tries to blackmail developers when they threaten to replace …
Anthropic’s new AI model turns to blackmail when engineers try to take it offline
https://techcrunch.com/2025/05/22/anthropics-new-ai-model-turns-to-blackmail-when-engineers-try-to-take-it-offline/
Tomi Engdahl says:
Sam Bowman / @sleepinyourhat:
Anthropic says Opus 4 will use an email tool to “whistleblow” if it detects users doing something “egregiously evil”, like marketing a drug based on faked data
https://x.com/sleepinyourhat/status/1925619851306918105
Tomi Engdahl says:
Eliot Brown / Wall Street Journal:
OpenAI partners with G42 and others to build a 1-gigawatt AI data center in Abu Dhabi called Stargate UAE, its first large-scale project outside the US — Stargate UAE project is part of a push by the Gulf petrostate to become a big player in artificial intelligence
OpenAI Commits to Giant U.A.E. Data Center in Global Expansion
Stargate UAE project is part of a push by the Gulf petrostate to become a big player in artificial intelligence
https://www.wsj.com/tech/open-ai-abu-dhabi-data-center-1c3e384d?st=tfh4w9&reflink=desktopwebshare_permalink
Tomi Engdahl says:
Hayden Field / CNBC:
Anthropic’s Jared Kaplan says the company stopped investing in chatbots at the end of 2024 and instead focused on improving Claude’s ability to do complex tasks
Anthropic launches Claude 4, its most powerful AI model yet
https://www.cnbc.com/2025/05/22/claude-4-opus-sonnet-anthropic.html
Tomi Engdahl says:
Richard Nieva / Forbes:
An interview with Niantic CEO John Hanke on the company selling its games business, pivoting to enterprise AI, returning to its digital mapping roots, and more
Pokemon Go Made Niantic Billions. Now It’s Ditching Gaming For AI.
https://www.forbes.com/sites/richardnieva/2025/05/21/niantic-scopely-pokemon-go/
Niantic was still raking in hundreds of millions a year from the beloved mobile game when CEO John Hanke decided to sell its games business and pivot to enterprise AI. Why blow it all up now?
In March, Niantic made a bombshell announcement: the developer of Pokemon Go — once the biggest mobile game ever in the U.S. — is abandoning games to go all-in on AI. It has sold off its game development business to Saudi-owned game maker Scopely in a $3.5 billion deal and rebranded itself as Niantic Spatial. Instead of building augmented reality games for mobile phones, it will develop artificial intelligence models that analyze the real world for enterprise clients.
“It’s kind of unusual for a successful company to do this cellular division — form two companies,” cofounder and CEO John Hanke told Forbes. “It became clear to us that the way to maximize the opportunity for both was to let each of them go and pursue its future.”
Now, Niantic is doubling down on its nascent Spatial platform, announced in November, which provides AI mapping tools that companies can use to chart out routes for robots or power augmented reality glasses. Just as large language models allow AI to generate text, Niantic’s Large Geospatial Models (LGMs) help AI understand, navigate and interact with physical spaces as a human would. The models are able to recreate 3D, real-world places thanks to Niantic’s massive set of location data, drawn from the 30 billion miles people have collectively walked playing its games like Pokemon Go and Ingress. And when the models don’t have precise data on all the dimensions, topography or physical structures in a place, they use generative AI to fill in those blanks, estimating different angles of a statue or missing corners of rooms.
“I don’t think maximizing the value for Pokemon Go for the next 10 years is necessarily where [Hanke’s] heart is at.”
Saar Gur, Partner, CRV
Niantic’s pivot underscores the seismic effect that the generative AI frenzy has had on Silicon Valley since ChatGPT rocked the industry nearly two-and-a-half years ago — radically transforming even a firmly established decade-old company like Niantic. According to Gartner, the market for spatial computing is expected to hit $1.7 trillion by 2033, up from $110 billion in 2023, with growth driven by location-based services from the likes of mapping giant TomTom and traditional big tech like Google. “The opportunity is enormous,” said Tuong Nguyen, director analyst for Garner’s emerging technology team.
So is the competition. In spatial AI, Niantic faces some formidable rivals. Since 2021, Nvidia, the $3 trillion chipmaker, has offered Omniverse, an enterprise platform that creates 3D “digital twins” for performing simulations in factories and other industrial settings. And last year, computer vision pioneer Fei-Fei Li, known as the Godmother of AI, founded World Labs, a startup building AI that generates 3D fantasy worlds, which could be helpful for video game development or astronaut simulations. The company is already valued at $1 billion — without even launching a product.
To fund its new company, Niantic went to its well of existing investors, including Coatue, Battery Ventures and CRV, for a $250 million investment. As part of the deal, which was in the works for a year and is expected to close by the end of the month, about 400 gaming employees will join Scopely, maker of the popular Monopoly Go mobile game, and about 200 will remain with Niantic.
From the start, Pokemon Go was a runaway hit, generating around $8 billion in revenue since its debut in 2016, analysts estimate. Almost a decade later, the game, which tasks players to catch virtual Pokemon by trekking to real-world locations, racked up 100 million players in 2024, Niantic said. The company brought in $1 billion in revenue last year, with 30 million monthly players across its catalog
Tomi Engdahl says:
Rhiannon Williams / MIT Technology Review:
Anthropic says Opus 4 marks a significant advancement in AI agents and that Rakuten deployed it to code autonomously for seven hours on a complicated project
Anthropic’s New Model Excels at Reasoning and Planning—and Has the Pokémon Skills to Prove It
Claude 4 Opus and Claude Sonnet 4 can remember over long periods of time—a capability that’s helpful at Pokémon and other tasks that require an ability to stay on track.
https://www.wired.com/story/anthropic-new-model-launch-claude-4/
Anthropic announced two new models, Claude 4 Opus and Claude Sonnet 4, during its first developer conference in San Francisco on Thursday. Claude 4 Opus will be immediately available to paying Claude subscribers, while Claude Sonnet 4 will be available to free and paid users.
Tomi Engdahl says:
Iltalehti voitti arvostetun kansainvälisen palkinnon – Tällainen on palkittu tekoälytyökalu
Iltalehti ja Alma Media on palkittu tekoälyn hyödyntämisestä arvostetussa kansainvälisessä media-alan kilpailussa.
https://www.iltalehti.fi/kotimaa/a/9d9af5a6-caf9-483c-a759-1efe95569b12
Iltalehti on voittanut ykköspalkinnon arvostetussa INMA Global Media Awardsissa. Palkinto myönnettiin Iltalehden tekoälyn avulla luomalle tunneluokitteluanalyysille, jolla taistellaan uutisväsymystä ja -välttelyä vastaan.
Iltalehti voitti Paras innovaatio uutistoimituksen muutoksessa (Best Innovation in Newsroom Transformation) -kategorian valtakunnallisten medioiden sarjan.
Tunneluokitteluanalyysin idea on tehdä toimitukselle näkyväksi se, millaisia tunteita yksittäiset sisällöt ja sisältötarjonta kokonaisuudessaan herättävät. Tämän avulla voidaan huolehtia siitä, että sisällöissä tarjolla on mahdollisimman tasapainoinen kokonaisuus positiivisia, neutraaleja ja negatiivisia tunteita. Tunneluokitteluanalyysi kertoo myös, kuinka vahvasti sisältö on esimerkiksi innostavaa, rakentavaa, surullista, viihdyttävää, empaattista, provosoivaa, analyyttistä, kyseenalaistavaa tai hyödyllistä.
Tomi Engdahl says:
Kuka tämän kirjoitti?
Tekoälyn kirjoitelmat on tärkeää voida erottaa ihmisen luomuksista. Tähän tarkoitettujen internetpalveluiden luotettavuus kuitenkin vaihtelee.
https://www.iltalehti.fi/oppaat/a/1d02ced6-53bb-4b61-bf06-e426e0425497
Tomi Engdahl says:
NVIDIA’s NVLink Spine Outpaces the Entire Internet—Welcome to the AI Superhighway
At Computex 2025, NVIDIA CEO Jensen Huang unveiled the NVLink Spine—a jaw-dropping leap in AI infrastructure. Boasting 130 terabytes per second of bandwidth across 5,000 fully meshed coaxial cables, this interconnect transfers more data than the entire global internet. Let that sink in.
This isn’t just GPU architecture. It’s a neural nexus—linking 72 GPUs into a single, ultra-coherent AI organism. Built for trillion-parameter models and next-gen AGI workloads, the NVLink Spine redefines scalability. It’s modular. It’s open. It’s even integrating with CPUs from Fujitsu and Qualcomm.
What the internet did for information, NVLink Fusion will do for intelligence. We’re not building computers anymore—we’re wiring sentience at scale. And with bandwidth like this, the future won’t just think faster. It’ll think together.
The AI era isn’t coming. It’s plugged in.
Media: @nvidia
https://www.facebook.com/share/p/1Dd9dM2bgD/
Tomi Engdahl says:
IBM laid off 8,000 employees to replace them with AI, but what they didn’t expect was having to rehire as many due to AI.
https://farmingdale-observer.com/2025/05/22/ibm-laid-off-8000-employees-to-replace-them-with-ai-but-what-they-didnt-expect-was-having-to-rehire-as-many-due-to-ai/?fbclid=IwY2xjawKeYJlleHRuA2FlbQIxMQABHgNt4SkL8GW96E-O9E0qyOzimQTF1NDFL-BkY5BDEHKuiqoJ1_tOz6u8IqTh_aem_vcHADMvblLN_b5RCVRhRyQ
In 2023, IBM made its mark by announcing the redundancy of almost 8,000 employees, mainly in human resources, to automate these functions using artificial intelligence. But against all odds, this strategy led to a wave of new hires… again because of AI.
In early 2023, IBM made a radical decision to replace thousands of jobs, mainly in support functions, with artificial intelligence solutions. The stated objective was clear: to automate up to 30% of repetitive tasks, particularly in human resources, and thus achieve significant productivity savings. This strategy, far from being isolated, was part of a worldwide wave of rationalisation in which other tech giants, such as Google and Spotify, were also making massive redundancies.
But IBM has taken its experimentation further than its competitors. Thanks to the AskHR solution, an AI-powered conversational agent, 94% of routine HR management tasks have been automated, from holiday processing to payroll management and employee documentation. This automation has enabled the firm to generate $3.5 billion in productivity across more than 70 different business lines.
However, what followed surprised even IBM’s top management. Far from reducing its workforce permanently, the company saw its total number of employees rise again after the wave of redundancies.
“While we’ve done a tremendous amount of work to leverage AI, our total employment has actually increased because it’s allowed us to invest more in other areas.”
IBM’s example illustrates a fundamental trend: AI is not just eliminating jobs, it is creating new ones, often with higher qualifications and better pay. But this transformation is not without its problems. The jobs most exposed to automation, such as support functions, are being replaced, while demand is exploding for profiles capable of managing, designing and selling AI solutions. IBM is not an isolated case. Other companies, such as Duolingo and certain customer service platforms, have also tried to massively replace their staff with chatbots, sometimes with mixed results, forcing them to rehire specialists to compensate for the limits of automation.
The success of automation at IBM is based on the company’s ability to reinvest the savings made in growth sectors. The case of AskHR is emblematic: in 2024, the platform handled more than 11.5 million interactions, with a customer satisfaction rate (NPS) that rose from -35 to +74 in just a few years. But despite these successes, only 6% of requests still require human intervention, proof that the transformation is not complete and that certain skills are still essential. The IBM case raises the question of the future of work in the age of AI. According to a report by the World Economic Forum, almost 92 million jobs could disappear by 2030 as a result of automation, but new professions are already emerging at high speed.
At IBM, AI has not only eliminated jobs: it has forced the company to rethink its HR strategy in depth and reinvent the contours of modern employment.
Tomi Engdahl says:
This AI read company emails and threatened to expose an engineer’s affair. https://link.ie.social/g1emAd
Tomi Engdahl says:
Simon Willison / Simon Willison’s Weblog:
Highlights from the system prompts of Claude Opus 4 and Claude Sonnet 4, including model safety, avoiding sycophancy, and not regurgitating copyrighted content — Anthropic publish most of the system prompts for their chat models as part of their release notes.
Highlights from the Claude 4 system prompt
https://simonwillison.net/2025/May/25/claude-4-system-prompt/
Anthropic publish most of the system prompts for their chat models as part of their release notes. They recently shared the new prompts for both Claude Opus 4 and Claude Sonnet 4. I enjoyed digging through the prompts, since they act as a sort of unofficial manual for how best to use these tools. Here are my highlights, including a dive into the leaked tool prompts that Anthropic didn’t publish themselves.
Reading these system prompts reminds me of the thing where any warning sign in the real world hints at somebody having done something extremely stupid in the past. A system prompt can often be interpreted as a detailed list of all of the things the model used to do before it was told not to do them.
Tomi Engdahl says:
Noam Scheiber / New York Times:
Some Amazon engineers say managers have increasingly pushed them to use AI over the past year, raising output goals and becoming less forgiving about deadlines — Pushed to use artificial intelligence, software developers at the e-commerce giant say they must work faster and have less time to think.
https://www.nytimes.com/2025/05/25/business/amazon-ai-coders.html?unlocked_article_code=1.J08.DkGW.5075LbaW7jrv
Tomi Engdahl says:
Yoolim Lee / Bloomberg:
OpenAI sets up a legal entity in South Korea and plans a Seoul office, its third in Asia, and says the country has the most paying ChatGPT users outside the US
OpenAI to Set Up Shop in South Korea to Spur Further Growth
https://www.bloomberg.com/news/articles/2025-05-26/openai-to-set-up-shop-in-south-korea-to-spur-further-growth
Tomi Engdahl says:
Jo Constantz / Bloomberg:
Tools like Microsoft Copilot are helping execs, managers, and staff quickly catch up on work after a vacation by triaging emails and summarizing Slack threads
https://www.bloomberg.com/news/articles/2025-05-22/ai-tools-are-ceo-s-new-fix-for-the-post-vacation-inbox-mess
Tomi Engdahl says:
Sean Heelan / Sean Heelan’s Blog:
A security researcher details how he discovered a zero-day vulnerability in the Linux kernel’s SMB implementation by analyzing the code using OpenAI’s o3 API — In this post I’ll show you how I found a zeroday vulnerability in the Linux kernel using OpenAI’s o3 model.
How I used o3 to find CVE-2025-37899, a remote zeroday vulnerability in the Linux kernel’s SMB implementation
https://sean.heelan.io/2025/05/22/how-i-used-o3-to-find-cve-2025-37899-a-remote-zeroday-vulnerability-in-the-linux-kernels-smb-implementation/
In this post I’ll show you how I found a zeroday vulnerability in the Linux kernel using OpenAI’s o3 model. I found the vulnerability with nothing more complicated than the o3 API – no scaffolding, no agentic frameworks, no tool use.
Recently I’ve been auditing ksmbd for vulnerabilities. ksmbd is “a linux kernel server which implements SMB3 protocol in kernel space for sharing files over network.“. I started this project specifically to take a break from LLM-related tool development but after the release of o3 I couldn’t resist using the bugs I had found in ksmbd as a quick benchmark of o3’s capabilities. In a future post I’ll discuss o3’s performance across all of those bugs, but here we’ll focus on how o3 found a zeroday vulnerability during my benchmarking. The vulnerability it found is CVE-2025-37899 (fix here), a use-after-free in the handler for the SMB ‘logoff’ command. Understanding the vulnerability requires reasoning about concurrent connections to the server, and how they may share various objects in specific circumstances. o3 was able to comprehend this and spot a location where a particular object that is not referenced counted is freed while still being accessible by another thread. As far as I’m aware, this is the first public discussion of a vulnerability of that nature being found by a LLM.
Tomi Engdahl says:
Will Oremus / Washington Post:
Grok’s tendency to repeat debunked claims raises doubts about Elon Musk’s goal of making it a reliable source of information in high-stakes areas like medicine — Grok has proved popular with X users. But a string of bizarre blunders has threatened to turn it into a punchline.
Tech Policy
How Elon Musk’s ‘truth-seeking’ chatbot lost its way
Grok has proved popular with X users. But a string of bizarre blunders has threatened to turn it into a punch line.
https://www.washingtonpost.com/technology/2025/05/24/grok-musk-ai/?pwapi_token=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJyZWFzb24iOiJnaWZ0IiwibmJmIjoxNzQ4MDU5MjAwLCJpc3MiOiJzdWJzY3JpcHRpb25zIiwiZXhwIjoxNzQ5NDQxNTk5LCJpYXQiOjE3NDgwNTkyMDAsImp0aSI6IjI2MDgwNzc3LTI3YjMtNDM1MC1iNTdmLWY3MjQzMGI0ZDkzZCIsInVybCI6Imh0dHBzOi8vd3d3Lndhc2hpbmd0b25wb3N0LmNvbS90ZWNobm9sb2d5LzIwMjUvMDUvMjQvZ3Jvay1tdXNrLWFpLyJ9.TmnwT5TFOYj4aYzew9e5sARUDRObN0aFVCbN-ZaRLMs
Tomi Engdahl says:
Matthias Bastian / The Decoder:
Google expands Veo 3 availability to 71 additional countries, and Gemini Pro subscribers now get a trial pack of 10 Veo 3 generations on the Gemini web app — About 100 hours after its initial launch, Google is opening up its AI video model Veo 3 to users in 71 additional countries.
Google expands access to Veo 3, its viral new video model, through the Gemini app
https://the-decoder.com/google-expands-access-to-veo-3-its-viral-new-video-model-through-the-gemini-app/