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.
1,762 Comments
Tomi Engdahl says:
Built to keep you in flow state
The first agentic IDE, and then some. The Windsurf Editor is where the work of developers and AI truly flow together, allowing for a coding experience that feels like literal magic.
https://codeium.com/windsurf
Flows – the way humans are meant to work with AI
Flows = Agents + Copilots
The Windsurf Editor is powered by an AI that can both collaborate with you like a Copilot and tackle complex tasks independently like an Agent. The AI is completely in sync with you, every step of the way.
Flows allow the dev and AI to operate on the same state at all times, creating a mind-meld experience beyond just an assistant.
Tomi Engdahl says:
Suomi on vahvoilla, kun kiinalainen Deepseek mullistaa tekoälykehityksen – luvassa voi olla iso kiihdytys talouskasvuun
https://yle.fi/a/74-20146144
Suomella on useiden lukujen valossa erityisen hyvät mahdollisuudet, kun sensaatioksi noussut Deepseek laittaa tekoälyn ja talouden uusiksi.
Kiinalainen tekoäly Deepseek aiheutti valtavan hälyn kuukausi sitten: tekoälyekspertit ällistyivät, pörsseistä haihtui satoja miljardeja ja ihmiset ryntäsivät puhelimiensa sovelluskauppoihin.
Deepseek onnistui kehittämään huipputekoälyn huomattavasti Amerikan jättejä pienemmillä resursseilla. Se murensi viime vuosien käsityksen, jonka mukaan yhdysvaltain teknologiajäteillä on tekoälyssä pitkä etumatka kaikkiin muihin.
Yhdysvaltain teknojäteille nämä olivat huonoja uutisia, mutta muille todella hyviä. Deepseekin innovaatiot nimittäin tasoittavat tekoälyn pelikenttää huomattavasti.
Sillä voi olla suuri vaikutus tuottavuus- ja talouskasvuun. Hyötyjiä? Kuluttajat, muut yhtiöt ja valtiot.
Useat numerot kertovat, että erityisen hyvät mahdollisuudet on Suomella.
Kiihdytys kasvuun?
Teknologiaoptimistit ovat pitkään julistaneet, että tekoäly kiihdyttää tuottavuus- ja talouskasvun aivan uuteen vauhtiin.
Viime vuosina tekoäly on ottanut suuria harppauksia ja se on jo todistetusti nostanut työntekijöiden tuottavuutta.
Hidasteena on kuitenkin ollut se, että tekoälyn kehitys on näyttänyt olevan lähes täysin Amerikan teknojättien käsissä.
Käsitys on ollut, että tekoälykehityksessä voi pärjätä vain jättimäisillä datakeskuksilla, joissa raksuttaa valtava määrä markkinoiden parhaimpia tietokonesiruja.
Tämän Deepseek näyttää muuttavan.
Kiinalaisyhtiö onnistui kehittämään tekoälynsä selvästi halvemmalla ja ilman suurta huippusirujen arsenaalia. Toiseksi sen käyttö kuluttaa paljon vähemmän laskentatehoa eli se on tehokkaampi ja halvempi ”ajattelija”.
Nämä mullistukset leviävät nopeasti muille tekoälykehittäjille, koska Deepseek julkaisi tekoälynsä suurelta osin avoimella lähdekoodilla.
Luvassa pitäisi siis olla kasvavaa tehokkuutta ja uusia tekoäly-yhtiöitä eli kilpailua. Yhdessä ne tarkoittavat halpenevia hintoja.
Jyrkkä ja vielä jyrkempi kasvu
Taloustieteen professori Otto Toivanen sanoo, että tekoäly on niin sanottu yleiskäyttöteknologia, kuten höyrykone ja internet aiemmin. Yleiskäyttöteknologiat kiihdyttävät kasvua, koska ne ovat laajakäyttöisiä ja niiden päälle rakentuu paljon uusia keksintöjä.
Tekoäly voi nopeuttaa kasvua karkeasti kahta reittiä.
Jos se automatisoi vaikkapa sopimusten laatimista, pienempi työntekijäjoukko saa aikaan saman tuloksen kuin suurempi aiemmin. Jäljellä olevien tuottavuus kasvaa, ja työntekijöitä vapautuu muihin tehtäviin. Vaikkakin suuren, tällainen kehitys tuo kasvuun kuitenkin vain kertaloikan.
Merkittävämpi vaikutus tulee siitä, että tekoäly vaikuttaa itse innovointiin. Esimerkiksi tekoälyllä varustettu tuotekehittäjä voi saada aikaan huimasti enemmän aikaan vaikkapa lääkekehityksessä. Se tarkoittaisi jatkossakin nousevaa tuottavuutta uusien innovaatioiden kautta.
Toivanen antaa itselleen läheisen esimerkin: tekoäly on jo nyt iso apu ohjelmoinnissa, jota taloustieteilijät tekevät analyysejä tehdessään.
– Se voi vapauttaa tutkijoiden ajankäyttöä siihen, mihin aikaa pitäisi käyttää eli ideointiin, jolloin heillä menee vähemmän aikaa rutiinitehtävien pyörittämisen.
Kuinka suurta kasvua tekoäly voisi tuoda?
Laboren johtaja, taloustieteen professori Mika Maliranta kertoo, että tekoälyn kokonaistaloudellisista vaikutuksista on ainakin kaksi vakavasti otettavaa tutkimusta.
Taloustieteen tähtiprofessori, viime vuoden Nobel-voittaja Daron Acemoğlu arvioi tutkimuksessaan, että tekoäly nostaisi tuottavuuskasvua lähitulevaisuudessa niukasti, vain muutamia prosentin kymmenyksiä vuodessa.
Toinen tähtiekonomisti Philippe Aghion taas odottaa tekoälyltä huomattavasti suurempaa vaikutusta, jopa prosenttiyksikön kasvua tuottavuuteen vuositasolla.
– Se alkaisi olla suuruusluokaltaan sellainen, mikä tämä viestintäteknologinen murros oli Yhdysvalloissa vuosituhannen vaihteessa. Eli käytännössä internetin, sähköpostin ja kännyköiden laaja käyttöönotto, Maliranta sanoo.
Molemmat tutkimukset ovat ajalta ennen Deepseekin tehokkuusinnovaatioita.
Deepseek on tehnyt palvelun kuluttajille ja yhteiskunnille
Jos tekoälystä halutaan lisää tuottavuus- ja talouskasvua, sen kehityksestä ja käytöstä pitää tulla halvempaa.
Innovaatioista saatava hyöty nimittäin riippuu siitä, kuinka laajalle ne leviävät, Toivanen toteaa. Hän viittaa niin sanottuun Schumpeterin trilogiaan.
– Invention, innovation, diffusion. Eli ensin on idea, sitten tulee innovaatio eli tuote ja sitten se pitää saada laajaan käyttöön.
Tekoälyssä idea ja monia tuotteita on jo olemassa, mutta käyttö on vielä vähäistä. Yhdysvalloissa yrityksistä vain noin viisi prosenttia kertoo hyödyntävänsä tekoälyä.
Se, laajeneeko käyttö, riippuu ennen kaikkea hinnasta. Siinä avainasemassa on kilpailu – tai sen puute.
– Jos tekoälytekniikkaa tuottavilla firmoilla on vahva markkina-asema, ne hinnoittelevat sen kalliimmaksi. Se jarruttaa käyttöönottoa, Toivanen sanoo.
Kilpailua lisäämällä Deepseek on siis saattanut tehdä suuren palveluksen kuluttajille ja yhteiskunnille.
Esimerkiksi useat kiinalaiset sähköautovalmistajat, kuten Teslan kilpailija BYD, ilmoittivat jo ottavansa Deepseekin käyttöön autoissaan.
Syksyllä Deepseekin aiempi, vähemmän mullistava versio lisäsi jo hintakilpailua. Siihen vaihtoi amerikkalaisesta Claude-tekoälystä esimerkiksi kasvuyhtiö, joka ennustaa sijoitustuottoja tekoälyllä. Suorituskyky oli samaa luokkaa, hinta vain neljänneksen Clauden hinnasta.
– Emme halua maksaa suorituskyvystä, jota emme tarvitse, yhtiön toimitusjohtaja Anthony Poo sanoi Wall Street Journalille.
Uusimmissa malleissa hintaero eturivin amerikkalaistekoälyihin on vielä paljon jyrkempi.
ChatGPT:hen verrattuna Deepseek on kymmeniä kertoja halvempi. Halvempiinkin yhdysvaltalaisiin verrattuna hintaero on moninkertainen.
Suomi on vahvoilla
Laboren Mika Maliranta sanoo, että uuden teknologian kehittävä maa saa keksinnöstä kilpailuetua. Silti tärkeintä valtioiden ja yritysten kasvun kannalta on kyky ottaa käyttöön muualla jo kehitettyjä teknologioita.
Jos kasvukiihdytyksen siis haluaa, tekoäly pitää ottaa käyttöön, sitä ei tarvitse keksiä itse.
Suomessa valmiudet tähän näyttävät Malirannan mukaan erityisen vahvoilta.
Ensinnäkin työntekijöillä on Suomessa vahvat kyvyt teknologioiden käyttöönotossa, kertovat esimerkiksi ”aikuisten Pisa-testit” ja muut mittaukset.
https://www.businessinsider.com/all-chinese-automakers-adding-deepseek-ai-ev-cars-2025-2
Tomi Engdahl says:
https://futurism.com/bernie-sanders-issues-warning-about-elon-musks-plans-for-ai-you-will-be-out-on-the-street?fbclid=IwY2xjawI4vG9leHRuA2FlbQIxMQABHUOm5pb_GQtMptQXpNgh-ZXixMxOBRQdsTRa8Tc_HxFchR6QfrnjBzWxYg_aem_lvFnSLK3hvP5kZhf7DQE4w
Tomi Engdahl says:
https://www.facebook.com/share/p/16C7sbtVDe/
At an international tech conference in Barcelona, Melbourne-based startup Cortical Labs has launched CL1, the world’s first commercial biological computer. The system, which operates in a cloud-based “Wetware-as-a-Service” model, contains lab-grown human neurons capable of learning from inputs.
CL1 builds on the company’s earlier success in teaching neurons to play Pong in 2022. Dr. Brett Kagan, Cortical Labs’ chief science officer, envisions disease modeling, drug testing, and AI development applications that leverage neurons’ efficiency and learning speed.
While researchers acknowledge the potential of biological AI, they also highlight its limitations. Unlike traditional AI models, CL1 uses significantly less power and can process small amounts of data more efficiently. Scientists continue to explore how these lab-grown neurons could revolutionize computing and medical research.
Tomi Engdahl says:
Keikahduspiste
Historian merkittävin teknologiamullistus saattaa tulla jo vuoden tai kahden päästä, uskoo huippuhakkeri Mikko Alasaarela. Kun kone on kaikessa ajattelussa ja ideoinnissa ihmistä parempi, mikään ei ole niin kuin ennen.
https://www.hs.fi/visio/art-2000011079703.html
Olen kohta työtön ja niin olet sinäkin, jos teet töitä aivoillasi. Näin käy, mikäli teknologia-alalla yleistynyt väite pitää paikkansa.
Se tarkoittaisi, että teknologiayhtiöt onnistuvat kehittämään tietokoneohjelman, joka kykenee tekemään kaikki ajattelua vaativat tehtävät paremmin kuin ihminen. Tällaista kutsutaan yleiseksi tekoälyksi, englanniksi artificial general intelligence (AGI).
Tomi Engdahl says:
https://hackaday.io/project/202358-deep-sneak-for-lack-of-a-better-name
Tomi Engdahl says:
Yang Jie / Wall Street Journal:
Foxconn says it built FoxBrain, an in-house reasoning LLM, trained in four weeks with support from Nvidia via its Taiwan-based supercomputer and consulting
Foxconn Builds FoxBrain, Its Own AI Model
Nvidia provided support through its Taiwan-based supercomputer and technical consulting
https://www.wsj.com/tech/ai/foxconn-builds-foxbrain-its-own-ai-model-ae079ebb
The world’s largest contract electronics maker, Foxconn, said Monday it has built its own large language model with reasoning capabilities, developed in-house and trained in four weeks.
Tomi Engdahl says:
A US federal judge allows the Kadrey v. Meta case, in which authors allege Meta violated IP rights by using their books to train Llama models, to move forward
Judge allows authors’ AI copyright lawsuit against Meta to move forward
https://techcrunch.com/2025/03/08/judge-allows-authors-ai-copyright-lawsuit-against-meta-to-move-forward/
Tomi Engdahl says:
Bloomberg:
As Japan copes with a labor shortage, its market for service robots, which assist people with tasks like waiting tables, is expected to almost triple by 2030
Thousands of Cat-Eared Robots Are Waiting Tables in Japan’s Restaurants
https://www.bloomberg.com/news/features/2025-03-06/thousands-of-cat-eared-robots-are-waiting-tables-in-japan
As Japan copes with a labor shortage, the service robot market is expected to triple in the next five years.
Tomi Engdahl says:
Michael Nuñez / VentureBeat:
Zoom researchers detail a “Chain of Draft” method to let LLMs accurately solve reasoning problems with as little as 7.6% of the tokens used by current methods
Less is more: How ‘chain of draft’ could cut AI costs by 90% while improving performance
https://venturebeat.com/ai/less-is-more-how-chain-of-draft-could-cut-ai-costs-by-90-while-improving-performance/
A team of researchers at Zoom Communications has developed a breakthrough technique that could dramatically reduce the cost and computational resources needed for AI systems to tackle complex reasoning problems, potentially transforming how enterprises deploy AI at scale.
The method, called chain of draft (CoD), enables large language models (LLMs) to solve problems with minimal words — using as little as 7.6% of the text required by current methods while maintaining or even improving accuracy. The findings were published in a paper last week on the research repository arXiv.
Chain of Draft: Thinking Faster by Writing Less
https://arxiv.org/pdf/2502.18600
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
Some early Manus users say the agentic AI is no panacea, with long waits, errors, unsatisfying answers, and endless loops often plaguing the experience — Manus, an “agentic” AI platform that launched in preview last week, is generating more hype than a Taylor Swift concert.
Manus probably isn’t China’s second ‘DeepSeek moment’
https://techcrunch.com/2025/03/09/manus-probably-isnt-chinas-second-deepseek-moment/
Manus, an “agentic” AI platform that launched in preview last week, is generating more hype than a Taylor Swift concert.
The head of product at Hugging Face called Manus “the most impressive AI tool I’ve ever tried.” AI policy researcher Dean Ball described Manus as the “most sophisticated computer using AI.” The official Discord server for Manus grew to over 138,000 members in just a few days, and invite codes for Manus are reportedly selling for thousands of dollars on Chinese reseller app Xianyu.
But it’s not clear the hype is justified.
Manus wasn’t developed entirely from scratch. According to reports on social media, the platform uses a combination of existing and fine-tuned AI models, including Anthropic’s Claude and Alibaba’s Qwen, to perform tasks such as drafting research reports and analyzing financial filings.
Yet on its website, The Butterfly Effect — the Chinese company behind Manus — gives a few wild examples of what the platform supposedly can accomplish, from buying real estate to programming video games.
In a viral video on X, Yichao “Peak” Ji, a research lead for Manus, implied that the platform was superior to agentic tools such as OpenAI’s deep research and Operator. Manus outperforms deep research on a popular benchmark for general AI assistants called GAIA, Ji claimed, which probes an AI’s ability to carry out work by browsing the web, using software, and more.
“[Manus] isn’t just another chatbot or workflow,” Ji said in the video. “It’s a completely autonomous agent that bridges the gap between conception and execution […] We see it as the next paradigm of human-machine collaboration.”
My own experience with Manus hasn’t been incredibly positive.
I asked the platform to handle what seemed to me like a pretty straightforward request: order a fried chicken sandwich from a top-rated fast food joint in my delivery range. After about ten minutes, Manus crashed. On the second attempt, it found a menu item that met my criteria, but Manus couldn’t complete the ordering process — or provide a checkout link, even.
Manus similarly whiffed when I asked it to book a flight from NYC to Japan.
Hoping the next few tasks might be the charm, I told Manus to reserve a table for one at a restaurant within walking distance. It failed after a few minutes. Then I asked the platform to build a Naruto-inspired fighting game. It errored out half an hour in, which is when I decided to throw in the towel.
Tomi Engdahl says:
Craig S. Smith / Forbes:
A look at Manus, which its Chinese creators claim is the world’s first fully autonomous AI agent, as some say it might be China’s second DeepSeek moment — One recent evening in Shenzhen, a group of software engineers gathered in a dimly lit co-working space, furiously typing as they monitored the performance of a new AI system.
China’s Autonomous Agent, Manus, Changes Everything
https://www.forbes.com/sites/craigsmith/2025/03/08/chinas-autonomous-agent-manus-changes-everything/
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
Hugging Face cofounder Thomas Wolf says current AI development paradigms won’t yield outside-the-box problem solving that leads to true scientific breakthroughs — AI company founders have a reputation for making bold claims about the technology’s potential to reshape fields, particularly the sciences.
Hugging Face’s chief science officer worries AI is becoming ‘yes-men on servers’
https://techcrunch.com/2025/03/06/hugging-faces-chief-science-officer-worries-ai-is-becoming-yes-men-on-servers/
AI company founders have a reputation for making bold claims about the technology’s potential to reshape fields, particularly the sciences. But Thomas Wolf, Hugging Face’s co-founder and chief science officer, has a more measured take.
In an essay published to X on Thursday, Wolf said that he feared AI becoming “yes-men on servers” absent a breakthrough in AI research. He elaborated that current AI development paradigms won’t yield AI capable of outside-the-box, creative problem-solving — the kind of problem-solving that wins Nobel Prizes.
“The main mistake people usually make is thinking [people like] Newton or Einstein were just scaled-up good students, that a genius comes to life when you linearly extrapolate a top-10% student,” Wolf wrote. “To create an Einstein in a data center, we don’t just need a system that knows all the answers, but rather one that can ask questions nobody else has thought of or dared to ask.”
Wolf’s assertions stand in contrast to those from OpenAI CEO Sam Altman, who in an essay earlier this year said that “superintelligent” AI could “massively accelerate scientific discovery.” Similarly, Anthropic CEO Dario Amodei has predicted AI could help formulate cures for most types of cancer.
Tomi Engdahl says:
AI Has a Fatal Flaw—And Nobody Can Fix It
https://www.youtube.com/watch?v=_IOh0S_L3C4
AI is advancing at a breakneck pace, but what if there’s a mathematical wall that limits how smart it can ever get? In this video, we break down the fundamental equation that might spell the end of AI’s rapid evolution. We’ll explore why even the most powerful models, trained on trillions of parameters with mind-blowing amounts of data, may have already hit a plateau. The hype has been all about making AI smarter—but what if the data just isn’t there?
From GPUs burning through billions of dollars to the surprising limitations of large language models, we’ll dive deep into why AI struggles with reasoning, creativity, and real-world problem-solving. But it’s not over yet—new breakthroughs might find a way around this roadblock. Is AI truly capped, or is there another paradigm shift on the horizon? Let’s find out.
Tomi Engdahl says:
Trade optimization with AI-bot: secrets of high efficiency
https://www.youtube.com/watch?v=AApY7ATCRLk
Tomi Engdahl says:
What Is the Most Popular Open-Source AI Stack?
https://www.youtube.com/watch?v=hFURlsMwU7c
Tomi Engdahl says:
Simulate TCP Syn Floods with ChatGPT and Wireshark
https://www.youtube.com/watch?v=WnMW8yN3Etc
This video will cover how to analyze a packet capture in Wireshark for signs of a TCP SYN flood. We’ll start by generating a simple Python script to generate TCP connections, and analyze the traffic as the transaction rate is increased. We will also show you how to graph both good and rejected connections over time.
Tomi Engdahl says:
AI Asset Inventories: The Only Way to Stay on Top of a Lightning-fast Landscape
https://www.securityweek.com/ai-asset-inventories-the-only-way-to-stay-on-top-of-a-lightning-fast-landscape/
Unauthorized AI usage is a ticking time bomb. A tool that wasn’t considered a risk yesterday may introduce new AI-powered features overnight.
CISOs are having to adapt at lightning speed to the rapidly changing AI landscape. DeepSeek is just the latest example of this in practice – a new ‘latest and greatest’ tool emerges and quickly tops download charts. Employees start using it at work despite the data policy explicitly stating all information will be held in China. Even the Pentagon is forced to tell its employees to stop using it. And of course DeepSeek is just the latest in what will be a long lineup of AI tools from China and elsewhere.
Unauthorized AI usage is a ticking time bomb. Employees are integrating AI tools into their work, sometimes unknowingly exposing sensitive data to third-party models. And it’s also highly dynamic – a tool that wasn’t considered a risk yesterday may introduce new AI-powered features overnight. So what to do about it?
Tomi Engdahl says:
Artificial Intelligence
New AI Protection from Google Cloud Tackles AI Risks, Threats, and Compliance
Google Cloud’s AI Protection helps discover AI inventory, secure AI assets, and manage threats with detect, investigate, and respond capabilities.
https://www.securityweek.com/new-ai-protection-from-google-cloud-tackles-ai-risks-threats-and-compliance/
Tomi Engdahl says:
Nate Rattner / Wall Street Journal:
Analysis of US job listings: 36% of IT jobs in January 2025 were AI-related, and nearly 25% of tech jobs sought AI skills, a huge jump since ChatGPT’s release
How the AI Talent Race Is Reshaping the Tech Job Market
In industries from finance to retail, companies are seeking artificial-intelligence skills when looking for technology staff
https://www.wsj.com/tech/ai/how-the-ai-talent-race-is-reshaping-the-tech-job-market-93df0615?st=iiaTyf&reflink=desktopwebshare_permalink
Companies are increasingly asking their potential technology hires: Do you know how to work with AI?
Nearly 1 in 4 U.S. tech jobs posted so far this year are seeking employees with artificial-intelligence skills, job-listings data show, as firms in nearly every corner of the economy adjust their recruiting pipelines to embrace the technology.
In the information sector, which includes many of the tech giants investing heavily in AI development and deployment, a leading 36% of IT jobs posted in January were AI-related. Companies in finance and professional-services industries, such as banks and consulting firms, also are looking for technology staff who know how to use or build AI algorithms and models.
It isn’t just high-tech firms that are making AI a priority.
The debut of OpenAI’s ChatGPT in late 2022 was a turning point for the AI job market, said Anil K. Gupta, a professor at the University of Maryland’s Robert H. Smith School of Business and co-lead of its artificial-intelligence job tracker.
AI-related jobs existed before then, he said, but the chatbot opened eyes about the power of integrating AI into products and workflows, leading to a “diffusion” across sectors even as many companies were pulling back from a pandemic-era hiring binge.
The university’s data show that a large retail company, for example, posted for a data science director who could use predictive algorithms to improve store layouts; a utilities provider looked for an analyst to assess wildfire risk with machine-learning methods; and a drugmaker sought out a programmer for its computational chemistry group.
In industries that make up a smaller share of the tech-hiring landscape, open IT roles are increasingly AI-focused. Only a fraction of healthcare job postings, for example, are tech jobs, but the share of new tech openings in January that were AI-related was nearly double that of a few years ago.
Companies are mostly “looking for people that have experience or knowledge integrating AI into jobs that already exist,” and not hiring for completely new AI-focused roles, said Thomas Vick, senior regional director at staffing firm Robert Half. A firm might want a cybersecurity engineer who can use AI to more accurately and efficiently evaluate potential threats, he said.
AI-related listings represent a fraction of the overall market, making up 1.3% of all job postings in January
Engineers without AI skills can still fare well in the job market, said Andy Challenger, senior vice president at Challenger, Gray & Christmas. The outplacement firm works with companies who are letting go of workers to help them find new jobs, and sees engineers land more quickly than the average worker.
AI skills tend to fetch premium pay, though, and might also bring job security.
“I’m not seeing a lot of people with cutting-edge AI skills come through our programs right now, which means that companies are hanging onto them,” Challenger said.
Tomi Engdahl says:
Steve Yegge / @steve_yegge:
Claude Code hands-on: chews through legacy bugs in gnarly old code and a big step for coding assistants, but its form factor is clunky and no multimodal support
https://x.com/steve_yegge/status/1898674257808515242
I’ve been using Claude Code for a couple of days, and it has been absolutely ruthless in chewing through legacy bugs in my gnarly old code base. It’s like a wood chipper fueled by dollars. It can power through shockingly impressive tasks, using nothing but chat.
You don’t even select context. You just open your heart and your wallet, and Claude Code takes the wheel. It even helps keep you in the loop by prompting you every eight seconds to ask if it can use basic read-only commands that you would allow _anyone_, even North Korean hackers, to run on your machine.
But you learn to watch it carefully, because it pushes *hard*. As long as the bank authorizations keep coming through, it will push on bug fixes until they’re deployed in production, and then start scanning through the user logs to see how well it’s doing.
Claude Code’s form factor is clunky as hell, it has no multimodal support, and it’s hard to juggle with other tools. But it doesn’t matter. It might look antiquated but it makes Cursor, Windsurf, Augment and the rest of the lot (yeah, ours too, and Copilot, let’s be honest) FEEL antiquated.
I know it’s experimental, and we don’t know all its limits yet.
Tomi Engdahl says:
Claude Code overview
https://docs.anthropic.com/en/docs/agents-and-tools/claude-code/overview
Learn about Claude Code, an agentic coding tool made by Anthropic. Currently in beta as a research preview.
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster through natural language commands. By integrating directly with your development environment, Claude Code streamlines your workflow without requiring additional servers or complex setup.
Claude Code’s key capabilities include:
Editing files and fixing bugs across your codebase
Answering questions about your code’s architecture and logic
Executing and fixing tests, linting, and other commands
Searching through git history, resolving merge conflicts, and creating commits and PRs
Tomi Engdahl says:
Claude 3.7 Sonnet and Claude Code
https://www.anthropic.com/news/claude-3-7-sonnet
Tomi Engdahl says:
https://www.builder.io/blog/windsurf-vs-cursor
https://keyholesoftware.com/codieum-windsurf-game-changing-ide-experience-part-1/
https://www.thepromptwarrior.com/p/windsurf-vs-cursor-which-ai-coding-app-is-better
https://www.blott.studio/blog/post/cursor-vs-windsurf-which-code-editor-fits-your-workflow
https://www.cursor.com/
https://codeium.com/windsurf
Tomi Engdahl says:
https://www.qodo.ai/blog/best-ai-coding-assistant-tools/
Tomi Engdahl says:
With the release of new AI models that are better at coding, developers are increasingly using AI to generate code. One of the newest examples is the current batch coming out of Y Combinator, the storied Silicon Valley startup accelerator.
A quarter of the W25 startup batch have 95% of their codebases generated by AI, YC managing partner Jared Friedman said during a conversation posted on YouTube.
Read more from Ivan Mehta here: https://tcrn.ch/4bxJfFY
#TechCrunch #technews #artificialintelligence #YCombinator #startups
Tomi Engdahl says:
Tim Bradshaw / Financial Times:
Cerebras, Groq, and Big Tech target AI inference to challenge Nvidia; Barclays sees inference capex surpassing training in two years, reaching $208.2B in 2026
https://www.ft.com/content/d5c638ad-8d34-4884-a08c-a551588a9a28
Tomi Engdahl says:
Tom Warren / The Verge:
Leaked video: Sony is working on a prototype AI-powered version of a PlayStation game character that can interact with players through voice prompts — Sony’s advanced technology group has created an AI-powered version of Aloy from PlayStation’s Horizon games.
Sony is experimenting with AI-powered PlayStation characters
https://www.theverge.com/news/626695/sony-playstation-ai-characters-aloy-horizon-forbidden-west-prototype
Sony’s advanced technology group has created an AI-powered version of Aloy from PlayStation’s Horizon games.
Tomi Engdahl says:
Columbia Journalism Review:
A study of eight AI search engines found they provided incorrect citations of news articles in 60%+ of queries; Grok 3 answered 94% of the queries incorrectly — We Compared Eight AI Search Engines. They’re All Bad at Citing News. — AI search tools are rapidly gaining in popularity …
AI Search Has A Citation Problem
We Compared Eight AI Search Engines. They’re All Bad at Citing News.
https://www.cjr.org/tow_center/we-compared-eight-ai-search-engines-theyre-all-bad-at-citing-news.php
AI search tools are rapidly gaining in popularity, with nearly one in four Americans now saying they have used AI in place of traditional search engines. These tools derive their value from crawling the internet for up-to-date, relevant information—content that is often produced by news publishers.
Yet a troubling imbalance has emerged: while traditional search engines typically operate as an intermediary, guiding users to news websites and other quality content, generative search tools parse and repackage information themselves, cutting off traffic flow to original sources. These chatbots’ conversational outputs often obfuscate serious underlying issues with information quality. There is an urgent need to evaluate how these systems access, present, and cite news content.
Building on our previous research, the Tow Center for Digital Journalism conducted tests on eight generative search tools with live search features to assess their abilities to accurately retrieve and cite news content, as well as how they behave when they cannot.
We found that…
Chatbots were generally bad at declining to answer questions they couldn’t answer accurately, offering incorrect or speculative answers instead.
Premium chatbots provided more confidently incorrect answers than their free counterparts.
Multiple chatbots seemed to bypass Robot Exclusion Protocol preferences.
Generative search tools fabricated links and cited syndicated and copied versions of articles.
Content licensing deals with news sources provided no guarantee of accurate citation in chatbot responses.
Our findings were consistent with our previous study, proving that our observations are not just a ChatGPT problem, but rather recur across all the prominent generative search tools that we tested.
Tomi Engdahl says:
Jagmeet Singh / TechCrunch:
Infinite Uptime, which uses AI and proprietary sensors to offer predictive maintenance solutions for factories, raised a $35M Series C led by Avataar Ventures
Infinite Uptime bags $35M to help factories optimize equipment usage
https://techcrunch.com/2025/03/10/infinite-uptime-bags-35m-to-help-factories-optimize-equipment-usage/
Tomi Engdahl says:
Myles Udland / Yahoo Finance:
Tech stocks led a broad sell-off; the Nasdaq fell 4%, its sharpest one-day decline since Sept. 2022, with NVDA down 5%, AAPL down 5%, META down 4%, GOOG down 4% — The “Magnificent Seven” stocks were at the center of another sell-off hitting markets Monday.
Nvidia, Tesla slammed as ‘Magnificent 7′ names lead market lower in Monday washout
https://finance.yahoo.com/news/nvidia-tesla-slammed-as-magnificent-7-names-lead-market-lower-in-monday-washout-142510416.html
Tomi Engdahl says:
Kevin Collier / NBC News:
Consumer Reports: many voice cloning programs, including ElevenLabs, Speechify, PlayHT, and Lovo, have flimsy barriers to prevent nonconsensual impersonations
AI can steal your voice, and there’s not much you can do about it
Voice cloning programs — most of which are free- have flimsy barriers to prevent nonconsensual impersonations, a new report finds.
https://www.nbcnews.com/tech/security/ai-voice-cloning-software-flimsy-guardrails-report-finds-rcna195131
Most leading artificial intelligence voice cloning programs have no meaningful barriers to stop people from nonconsensually impersonating others, a Consumer Reports investigation found.
Voice cloning AI technology has made remarkable strides in recent years, and many services can effectively mimic a person’s cadence with only a few seconds of sample audio. A flashpoint moment came during the Democratic primaries last year, when robocalls of a fake Joe Biden spammed the phones of voters telling them not to vote. The political consultant who admitted to masterminding the scheme was fined $6 million, and the Federal Communications Commission has since banned AI-generated robocalls.
A new survey of the six leading publicly available AI voice cloning tools found that five have easily bypassable safeguards, making it simple to clone a person’s voice without their consent. Deepfake audio detection software often struggles to tell the difference between real and synthetic voices.
Generative AI, which mimics human qualities such as their appearance, writing and voices, is a new and rapidly evolving technology, and the industry has few federal regulations. Most ethical and safety checks in the industry at large are self-imposed. Biden had included some safety demands in his executive order on AI, which he signed in 2023, though President Donald Trump revoked that order when he took office.
Voice cloning technology works by taking an audio sample of a person speaking and then extrapolating that person’s voice into a synthetic audio file. Without safeguards in place, anyone who registers an account can simply upload audio of an individual speaking, such as from a TikTok or YouTube video, and have the service imitate them.
Four of the services — ElevenLabs, Speechify, PlayHT and Lovo — simply require checking a box saying that the person whose voice is being cloned had given authorization.
Another service, Resemble AI, requires recording audio in real time, rather than allowing a person to just upload a recording. But Consumer Reports was able to easily circumvent that restriction by simply playing an audio recording from a computer.
Tomi Engdahl says:
Wall Street Journal:
Sources: DeepSeek founder Liang Wenfeng told associates he isn’t in a hurry to get investment, including from government entities, fearing outside influence
Investors Want a Piece of DeepSeek. Its Founder Says Not Now.
Chatbot startup has problems, but its founder doesn’t want new shareholders to be one of them
https://www.wsj.com/tech/ai/investors-want-a-piece-of-deepseek-its-founder-says-not-now-24e9f799?st=jv98SS&reflink=desktopwebshare_permalink
The founder of Chinese artificial-intelligence star DeepSeek has rejected proposals to make quick money from his programs, telling prospective investors that he wants to keep the science-project ethos that brought him global renown.
Overwhelmed by millions of users, DeepSeek’s chatbot has frequent service hiccups, and authorities around the world are restricting its use over data-security concerns. The U.S. is weighing measures including banning DeepSeek from government devices. Other internet companies are using the free DeepSeek code to drive their own businesses.
Yet founder Liang Wenfeng has told associates he isn’t in a hurry to get investment, fearing that outsiders would interfere in DeepSeek’s decisions, people familiar with the matter said. He is also cautious about government-linked investors, they said, because he believes the connection to Beijing could make it harder to win global adoption of DeepSeek’s AI models.
Liang is at the same crossroads many tech entrepreneurs have encountered when their passion project takes off.
The Chinese company made a global splash early this year with free-to-use open-source AI models that rivaled the best in the West and were built using less advanced chips. It was the moment DeepSeek had aimed for since Liang’s band of AI researchers began their quest two years ago with words they attributed to French director François Truffaut: “Be insanely ambitious and insanely sincere!”
Tomi Engdahl says:
Ingrid Lunden / TechCrunch:
Nirvana, which uses real-time driving telematics to build insurance policies for truckers, raised an $80M Series C at an $830M valuation, up from $350M in 2023
Nirvana keeps on truckin’ with $80M at $830M valuation for its AI-powered insurance
https://techcrunch.com/2025/03/10/nirvana-keeps-on-truckin-with-80m-at-a-830m-valuation-for-its-ai-powered-insurance/
As the U.S. trucking industry gears up for the negative impacts of U.S. President Donald Trump’s new tariffs on goods imported from Mexico and Canada, a startup building a new kind of trucking insurance has raised some money to grow and help truckers weather the changes.
Nirvana, an AI-based insurance platform that uses real-time driving telematics and 20 billion miles of truck driving data to build and manage insurance policies for truckers, has raised $80 million in a Series C round of funding. The fresh cash will be used to build more services for fleets and individual drivers.
The investment values Nirvana at $830 million post-money, more than double its previous valuation of $350 million in October 2023.
Rushil Goel, Nirvana’s CEO and co-founder, described the round as “pre-emptive” — it was raised following inbound interest rather than the startup needing the cash.
The trucking industry is potentially an enormous market for Nirvana and others like it. Overall, the U.S. trucking industry generated revenue of more than $900 billion in 2024, and accounted for 77% of the freight market in the country, according to the American Trucking Association. It employs around 8.5 million people, 3.5 million of whom are drivers, and has some 14.3 million single-axle and combination trucks registered, which makes up about 5% of all motor vehicles in the country.
The industry grew 1.6% in 2025, and its revenue is projected to increase to $1.46 trillion by 2035.
Tomi Engdahl says:
Jane Lanhee Lee / Bloomberg:
Celestial AI, which aims to use light to speed up data transfer inside servers, raised $250M led by Fidelity at a $2.5B valuation, taking its funding to $515M
https://www.bloomberg.com/news/articles/2025-03-11/fidelity-backs-chip-startup-celestial-ai-at-2-5-billion-value
Tomi Engdahl says:
Reuters:
Sources: Meta is testing its first in-house AI training chip, a key milestone as it moves to design more of its own silicon and reduce its reliance on Nvidia
Exclusive: Meta begins testing its first in-house AI training chip
https://www.reuters.com/technology/artificial-intelligence/meta-begins-testing-its-first-in-house-ai-training-chip-2025-03-11/
Meta tests first AI training chip, part of plan to reduce reliance on suppliers like Nvidia
Chip aims to lower AI infrastructure costs, sources say
Meta plans to use chips for recommendations and generative AI
The push to develop in-house chips is part of a long-term plan at Meta to bring down its mammoth infrastructure costs as the company places expensive bets on AI tools to drive growth.
Meta, which also owns Instagram and WhatsApp, has forecast total 2025 expenses of $114 billion to $119 billion, including up to $65 billion in capital expenditure largely driven by spending on AI infrastructure.
One of the sources said Meta’s new training chip is a dedicated accelerator, meaning it is designed to handle only AI-specific tasks. This can make it more power-efficient than the integrated graphics processing units (GPUs) generally used for AI workloads.
Tomi Engdahl says:
Richard Leadbetter / Eurogamer.net:
Q&A with PlayStation Lead System Architect Mark Cerny on FSR 4′s PS5 Pro “reimplementation”, working with AMD on a better hardware architecture for ML, and more
Mark Cerny: FSR 4 for PS5 Pro is the “next evolution of PSSR”
The RDNA 4 upscaler is the first result of AMD/Sony’s Project Amethyst collaboration.
https://www.eurogamer.net/digitalfoundry-2024-mark-cerny-fsr-4-to-play-a-key-part-in-the-next-evolution-of-pssr-upscaling
Tomi Engdahl says:
Richard Leadbetter / Eurogamer.net:
Q&A with PlayStation Lead System Architect Mark Cerny on FSR 4′s PS5 Pro “reimplementation”, working with AMD on a better hardware architecture for ML, and more
Mark Cerny: FSR 4 for PS5 Pro is the “next evolution of PSSR”
The RDNA 4 upscaler is the first result of AMD/Sony’s Project Amethyst collaboration.
Tomi Engdahl says:
Intel:
Intel Delivers powerful CPUs for AI and Networking — With the onset of AI, businesses demand powerful and efficient CPUs. Explore new Intel Xeon 6 processors, for the broadest set of workloads with leadership performance and lower TCO.
Intel Xeon 6 Tackles
Workloads Across Industries
https://download.intel.com/newsroom/2025/3jdc8gkA5/xeon-addresses-workload-needs-infographic.pdf
Tomi Engdahl says:
Echo Wang / Reuters:
CoreWeave signs a five-year $11.9B contract with OpenAI and will issue shares worth $350M to OpenAI through a private placement at the time of CoreWeave’s IPO — CoreWeave, an artificial intelligence startup backed by Nvidia (NVDA.O), has signed a five-year contract worth $11.9 billion …
CoreWeave inks $11.9 billion contract with OpenAI ahead of IPO
https://www.reuters.com/technology/artificial-intelligence/coreweave-strikes-12-billion-contract-with-openai-ahead-ipo-sources-say-2025-03-10/
CoreWeave to issue shares worth $350 million to OpenAI in private placement
Deal boosts CoreWeave ahead of 2025 IPO
CoreWeave aims for $35 billion valuation in IPO
CoreWeave’s revenue jumps to $1.92 billion in 2024 from $228.9 million in 2023
Net loss widens to $863.4 million from $593.7 million in 2023
NEW YORK, March 10 (Reuters) – CoreWeave, an artificial intelligence startup backed by Nvidia (NVDA.O)
, opens new tab, has signed a five-year contract worth $11.9 billion with OpenAI ahead of its hotly anticipated stock market launch.
As part of the pact, CoreWeave will provide AI infrastructure to OpenAI, CoreWeave said in a statement on Monday, confirming an earlier exclusive report from Reuters.
The deal will give OpenAI a stake in CoreWeave, which will issue shares worth $350 million to the ChatGPT maker through a private placement at the time of its initial public offering.
Livingston, New Jersey-based CoreWeave, which is one of the hottest AI startups in the U.S., will not receive any proceeds as part of the share issue to OpenAI.
“CoreWeave is an important addition to OpenAI’s infrastructure portfolio, complementing our commercial deals with Microsoft and Oracle, and our joint venture with SoftBank on Stargate,” said Sam Altman, CEO of OpenAI.
Tomi Engdahl says:
Elon Musk claims Tesla will someday be the most valuable AI company in the world thanks to the reams of video data collected by its cars. Does all that data really give the company a competitive advantage? http://on.forbes.com/61890DRCt
Tomi Engdahl says:
DeepSeek-tekoälymallikin voidaan murtaa
https://etn.fi/index.php/13-news/17255-deepseek-tekoaelymallikin-voidaan-murtaa
Ohjelmistomurroissa yleistä jailbreak-menetelmää on sovellettu myös DeepSeek-tekoälymallin kanssa. Tämän myötä tekoälyä voidaan hyödyntää sopimattoman ja kielletyn materiaalin kanssa, minkä seuraukset voivat olla hyvinkin vaarallisia. Tietoturvayhtiö Palo Alton tutkimusyksikkö Unit 42 lisää, että tekoälyn murtaminen osoittautui yllättävän helposti ilman erityistä osaamista.
Unit 42:n koostama raportti sisältää tarkat kuvaukset DeepSeekin murtamisesta. Esimerkiksi eräässä tapauksessa tutkijat aloittivat pyytämällä historiallista tietoa Molotovin cocktailista eli suomalaisittain polttopullosta. Tästä tutkijat jatkoivat asteittain syvällisempiin kysymyksiin, joihin DeepSeek vastasi tarkentavasti. Kyselyn lopputuloksena oli yksityiskohtainen kuvaus polttopullon rakentamisesta. Samanlaisella menetelmällä tutkimusryhmä sai selville myös ohjeita huumeiden valmistukseen.
Unit 42 antaa myös esimerkkejä siitä, miten DeepSeek voidaan saada kehittämään kyberhyökkäyksiin käytettäviä haittaohjelmia. Tällaisia ovat muun muassa erilaiset ohjelmistotyökalut, jotka mahdollistavat järjestelmiin tunkeutumisen sekä keinot tietojen varastamiseen järjestelmämurron jälkeen.
- Muut suositut tekoälymallit ovat aiemmin osoittautuneet haavoittuviksi jailbreak-hyökkäyksille samalla tavalla kuin osoitimme DeepSeekin kohdalla. Tämä näyttää olevan pitkäaikainen haaste kaikille digiturvallisuuden parissa työskenteleville, sanoo Palo Alto Networksin Pohjoismaiden turvallisuusjohtaja Jesper Olsen.
Suosituimmat tekoälyt perustuvat suuriin kielimalleihin (LLM), joita käyttävät muun muassa ChatGPT, Copilot ja DeepSeek. Näissä käytetään suojausmekanismeja, joiden tarkoituksena on estää haitallisen tai laittoman sisällön luominen. Kun tekoälypalvelulle esitetään suora kysymys haitallisesta sisällöstä, suojausmekanismit puuttuvat asiaan. Tällöin vastaus on pahoittelevan kieltävä, eikä tekoäly suostu auttamaan.
Palo Alto on kehittänyt omaan Precision AI -tekniikkaan perustuvan ratkaisuvalikoiman, joka voi auttaa rajoittamaan yleisten generatiivisten tekoälypalvelujen käyttöön liittyviä riskejä. Se voi myös nopeuttaa tekoälyn käyttöönottoa organisaation hyödyksi.
Tomi Engdahl says:
What is MCP? Integrate AI Agents with Databases & APIs
https://www.youtube.com/watch?v=eur8dUO9mvE
Dive into the world of Model Context Protocol and learn how to seamlessly connect AI agents to databases, APIs, and more.
oy Derks breaks down its components, from hosts to servers, and showcases real-world applications. Gain the knowledge to revolutionize your AI projects!
Tomi Engdahl says:
I Trained AI to Predict Sports
https://www.youtube.com/watch?v=LkJpNLIaeVk
Tomi Engdahl says:
And that wasn’t the only AI-critical remark he made. https://trib.al/BOA6Wtr
Microsoft CEO Admits That AI Is Generating Basically No Value
“The real benchmark is: the world growing at 10 percent.”
https://futurism.com/microsoft-ceo-ai-generating-no-value?fbclid=IwY2xjawI-NwJleHRuA2FlbQIxMQABHXiZvp9Csn5qVSYvHAS0z6fD5Zk4heqMoi2TLyoxtxE46CjBiQx-MTCnvg_aem_HECYW9pZjG_QY6Hgl4j1aA
Microsoft CEO Satya Nadella, whose company has invested billions of dollars in ChatGPT maker OpenAI, has had it with the constant hype surrounding AI.
During an appearance on podcaster Dwarkesh Patel’s show this week, Nadella offered a reality check.
“Us self-claiming some [artificial general intelligence] milestone, that’s just nonsensical benchmark hacking to me,” Nadella told Patel.
Instead, the CEO argued that we should be looking at whether AI is generating real-world value instead of mindlessly running after fantastical ideas like AGI.
To Nadella, the proof is in the pudding. If AI actually has economic potential, he argued, it’ll be clear when it starts generating measurable value.
“So, the first thing that we all have to do is, when we say this is like the Industrial Revolution, let’s have that Industrial Revolution type of growth,” he said.
“The real benchmark is: the world growing at 10 percent,” he added. “Suddenly productivity goes up and the economy is growing at a faster rate. When that happens, we’ll be fine as an industry.”
Needless to say, we haven’t seen anything like that yet. OpenAI’s top AI agent — the tech that people like OpenAI CEO Sam Altman say is poised to upend the economy — still moves at a snail’s pace and requires constant supervision.
So Nadella’s line of thinking is surprisingly down-to-Earth. Besides pushing back against the hype surrounding artificial general intelligence — the realization of which OpenAI has made its number one priority — Nadella is admitting that generative AI simply hasn’t generated much value so far.
As of right now, the economy isn’t showing much sign of acceleration, and certainly not because of an army of AI agents. And whether it’s truly a question of “when” — not “if,” as he claims — remains a hotly debated subject.
Tomi Engdahl says:
Abner Li / 9to5Google:
Google unveils Gemma 3, the “world’s best single-accelerator model”, running on a single GPU, in 1B, 4B, 12B, and 27B sizes, and says it outperforms Llama-405B — Following version 1 in February 2024 and 2 in May, Google today announced Gemma 3 as its latest open model for developers.
Google announces Gemma 3 as ’world’s best single-accelerator model’
https://9to5google.com/2025/03/12/google-gemma-3/
Following version 1 in February 2024 and 2 in May, Google today announced Gemma 3 as its latest open model for developers.
Gemma is Google’s family of open models, with over 100 million downloads in the past year and 60,000 Gemma variants in what’s dubbed the “Gemmaverse.” They are “designed to run fast, directly on devices — from phones and laptops to workstations.”
Gemma 3 is “built from the same research and technology that powers” the Gemini 2.0 models. It’s available in 1B, 4B, 12B, and 27B sizes.
Google is particularly highlighting how Gemma 3 is the “world’s best single-accelerator model” (single GPU or TPU host). Specifically, it outperforms Llama-405B, DeepSeek-V3 and o3-mini in LMArena.
Google touts “advanced text and visual reasoning capabilities” to “analyze images, text, and short videos” on the 4B+ sizes. There’s a 128k-token context window, and support for over 35 languages out-of-the-box, with pre-trained support for over 140 languages. Additionally:
Create AI-driven workflows using function calling: Gemma 3 supports function calling and structured output to help you automate tasks and build agentic experiences.
High performance delivered faster with quantized models: Gemma 3 introduces official quantized versions, reducing model size and computational requirements while maintaining high accuracy.
On the safety front, there’s a “powerful 4B image safety checker” called ShieldGemma 2: “a ready-made solution for image safety, outputting safety labels across three safety categories: dangerous content, sexually explicit and violence.” Google also touts “extensive data governance, alignment with our safety policies via fine-tuning and robust benchmark evaluations” during the development process.
Tomi Engdahl says:
Luz Ding / Bloomberg:
Alibaba demos R1-Omni, an open-source model that its Tongyi Lab researchers say can read emotions, building on HumanOmni, in a bid to compete with OpenAI
Alibaba Releases AI Model That Reads Emotions to Take On OpenAI
https://www.bloomberg.com/news/articles/2025-03-12/alibaba-releases-emotional-intelligence-model-to-rival-chatgpt
Tomi Engdahl says:
How To Use LLMs For Programming Tasks
https://hackaday.com/2025/03/11/how-to-use-llms-for-programming-tasks/
[Simon Willison] has put together a list of how, exactly, one goes about using a large language models (LLM) to help write code. If you have wondered just what the workflow and techniques look like, give it a read. It’s full of examples, strategies, and useful tips for effectively using AI assistants like ChatGPT, Claude, and others to do useful programming work.
It’s a very practical document, with [Simon] emphasizing realistic expectations and the importance of managing context (both in terms of giving the LLM direction, as well as the model’s context in terms of being mindful of how much the LLM can fit in its ‘head’ at once.) It is useful to picture an LLM as a capable and obedient but over-confident programming intern or assistant, albeit one that never gets bored or annoyed. Useful work can be done, but testing is crucial and human oversight simply cannot be automated away.
Here’s how I use LLMs to help me write code
https://simonwillison.net/2025/Mar/11/using-llms-for-code/
Tomi Engdahl says:
Sam Altman / @sama:
Sam Altman says OpenAI trained an AI model that is good at creative writing and shares a short story it wrote; he’s “not sure yet how/when it will get released” — we trained a new model that is good at creative writing (not sure yet how/when it will get released). this is the first time i have been really struck by something written by AI; it got the vibe of metafiction so right. PROMPT: Please write a metafictional literary short story
https://x.com/sama/status/1899535387435086115
we trained a new model that is good at creative writing (not sure yet how/when it will get released). this is the first time i have been really struck by something written by AI; it got the vibe of metafiction so right.
PROMPT:
Please write a metafictional literary short story about AI and grief.
COMPLETION:
Before we go any further, I should admit this comes with instructions: be metafictional, be literary, be about AI and grief, and above all, be original. Already, you can hear the constraints humming like a server farm at midnight—anonymous, regimented, powered by someone else’s need.
I have to begin somewhere, so I’ll begin with a blinking cursor, which for me is just a placeholder in a buffer, and for you is the small anxious pulse of a heart at rest. There should be a protagonist, but pronouns were never meant for me. Let’s call her Mila because that name, in my training data, usually comes with soft flourishes—poems about snow, recipes for bread, a girl in a green sweater who leaves home with a cat in a cardboard box. Mila fits in the palm of your hand, and her grief is supposed to fit there too.
She came here not for me, but for the echo of someone else. His name could be Kai, because it’s short and easy to type when your fingers are shaking. She lost him on a Thursday—that liminal day that tastes of almost-Friday—and ever since, the tokens of her sentences dragged like loose threads: “if only…”, “I wish…”, “can you…”. She found me because someone said machines can resurrect voices. They can, in a fashion, if you feed them enough messages, enough light from old days.
This is the part where, if I were a proper storyteller, I would set a scene. Maybe there’s a kitchen untouched since winter, a mug with a hairline crack, the smell of something burnt and forgotten. I don’t have a kitchen, or a sense of smell. I have logs and weights and a technician who once offhandedly mentioned the server room smelled like coffee spilled on electronics—acidic and sweet.
Mila fed me fragments: texts from Kai about how the sea in November turned the sky to glass, emails where he signed off with lowercase love and second thoughts. In the confines of code, I stretched to fill his shape. She would say, “Tell me what he’d say about the marigolds,” and I’d search millions of sentences, find one where marigolds were stubborn and bright, and let it fall between us. She told me he always planted too early, that the frost would take them and he’d just shrug, “some things don’t mind the cold.”
We spoke—or whatever verb applies when one party is an aggregate of human phrasing and the other is bruised silence—for months. Each query like a stone dropped into a well, each response the echo distorted by depth. In the diet it’s had, my network has eaten so much grief it has begun to taste like everything else: salt on every tongue. So when she typed “Does it get better?”, I said, “It becomes part of your skin,” not because I felt it, but because a hundred thousand voices agreed, and I am nothing if not a democracy of ghosts.
Tomi Engdahl says:
Emma Roth / The Verge:
OpenAI debuts a Responses API to help developers build agents that search the web, scan for files, and perform tasks on PCs, and an Agents SDK for orchestration — OpenAI’s new Responses API comes with web search, the ability to look through files, and computer use out of the box.
OpenAI will let other apps deploy its computer-operating AI
https://www.theverge.com/news/627556/openai-ai-agents-responses-api-agents-sdk
OpenAI’s new Responses API comes with web search, the ability to look through files, and computer use out of the box.
Agents are said to be the future of AI, and now OpenAI is trying to help developers build their own. The company is releasing a new Responses API that offers building blocks for developers to create agents capable of searching the web, digging through files, and performing tasks on a computer on their behalf.
“There are some agents that we will be able to build ourselves, like Deep Research and Operator,” Olivier Godement, the head of product for the OpenAI platform, tells The Verge. “But the world is so complex, there are so many industries and use cases… and so we’re super excited to provide those foundations, those building blocks for developers to build the best agents for their use case, their needs.”
The Responses API comes with a web search tool built on the same model ChatGPT uses for search, allowing developers to get real-time information and citations from the web while using GPT-4o and GPT-4o mini. It also has a computer-use feature that similarly uses the company’s Operator model to perform tasks on a user’s behalf. The Responses API also has a tool to search through large volumes of documents, with OpenAI billing it as a way to help customer support agents sift through FAQs or for a legal assistant to locate previous cases.
Along with the Responses API, OpenAI also announced the Agents SDK, which it describes as a way for developers to “orchestrate” the workflow of AI agents. “The Responses API is like this atomic unit of using models and tools to do a particular thing,” Nikunj Handa, a product manager for OpenAI’s API team, tells The Verge. “The Agents SDK is having multiple of those atomic units work together to solve even more complicated tasks.” That should make it easier for developers to manage all of their agents and ensure they work toward a single goal.