Artificial intelligence is rapidly changing many aspects of how we work and live. (How many stories did you read last week about self-driving cars and job-stealing robots? Perhaps your holiday shopping involved some AI algorithms, as well.) But despite the constant flow of news, many misconceptions about AI remain.
AI doesn’t think in our sense of the word at all, Scriffignano explains. “In many ways, it’s not really intelligence. It’s regressive.”
IT leaders should make deliberate choices about what AI can and can’t do on its own. “You have to pay attention to giving AI autonomy intentionally and not by accident,”
6,747 Comments
Tomi Engdahl says:
PLCHound Aims to Improve Detection of Internet-Exposed ICS
Georgia Tech researchers have developed PLCHound, an algorithm that uses AI to improve the identification of internet-exposed ICS.
https://www.securityweek.com/plchound-aims-to-improve-detection-of-internet-exposed-ics/
Tomi Engdahl says:
Financial Times:
Amazon’s Annapurna Labs plans to unveil “Trainium 2” AI chips next month, in an effort that can rival Nvidia; Trainium 2 is being tested by Anthropic and others
https://www.ft.com/content/3d9b5c6d-f1ae-4f6f-adc3-51e5f1dfb008
Tomi Engdahl says:
Reuters:
Ilya Sutskever says “we’re back in the age of wonder and discovery” as AI companies focus on pre-training, inference improvements, and finding the “next thing” — Artificial intelligence companies like OpenAI are seeking to overcome unexpected delays and challenges …
OpenAI and others seek new path to smarter AI as current methods hit limitations
https://www.reuters.com/technology/artificial-intelligence/openai-rivals-seek-new-path-smarter-ai-current-methods-hit-limitations-2024-11-11/
AI companies face delays and challenges with training new large language models
Some researchers are focusing on more time for inference in new models
Shift could impact AI arms race for resources like chips and energy
Nov 11 (Reuters) – Artificial intelligence companies like OpenAI are seeking to overcome unexpected delays and challenges in the pursuit of ever-bigger large language models by developing training techniques that use more human-like ways for algorithms to “think”.
A dozen AI scientists, researchers and investors told Reuters they believe that these techniques, which are behind OpenAI’s recently released o1 model, could reshape the AI arms race, and have implications for the types of resources that AI companies have an insatiable demand for, from energy to types of chips.
OpenAI declined to comment for this story. After the release of the viral ChatGPT chatbot two years ago, technology companies, whose valuations have benefited greatly from the AI boom, have publicly maintained that “scaling up” current models through adding more data and computing power will consistently lead to improved AI models.
But now, some of the most prominent AI scientists are speaking out on the limitations of this “bigger is better” philosophy.
Ilya Sutskever, co-founder of AI labs Safe Superintelligence (SSI) and OpenAI, told Reuters recently that results from scaling up pre-training – the phase of training an AI model that use s a vast amount of unlabeled data to understand language patterns and structures – have plateaued.
Sutskever is widely credited as an early advocate of achieving massive leaps in generative AI advancement through t he use of more data and computing power in pre-training, which eventually created ChatGPT. Sutskever left OpenAI earlier this year to found SSI.
Tomi Engdahl says:
Bloomberg:
11x, which makes AI bots that help salespeople with common tasks, raised $50M led by a16z at a $320M valuation, after raising a $24M Series A in September 2024
https://www.bloomberg.com/news/articles/2024-11-11/andreessen-horowitz-leads-50-million-funding-for-ai-startup-11x
Tomi Engdahl says:
7 vinkkiä: Näin hyödynnät tekoälyä sijoittamisessa
https://www.op-media.fi/blogit/tekoaly-sijoittamisen-apuna–7-vinkkia/
OP Uusimaan tutkimuksessa 35 % sijoittajista suhtautuu myönteisesti tekoälyn käyttöön sijoittamisessa. Vain 2 % on ottanut sen käyttöön. Nuoret ovat edellä: 16–25-vuotiaista yli puolet aikoo käyttää tekoälyä vaurastumisessa.
1. Hyödynnä tekoälyä datan yhdistelyssä ja analysoinnissa
Tekoäly analysoi valtavia tietomääriä nopeasti ja auttaa ymmärtämään markkinoita syvällisemmin.
2. Ennusta markkinatrendejä
Tekoälymallit voivat auttaa arvioimaan tulevia markkinaliikkeitä analysoimalla historiallista kurssidataa ja löytämällä toistuvia kaavoja suurista määristä erilaista numeerista tai kirjallista materiaalia. Sijoittajan tekniseen analyysiin tekoäly voi antaa lisätukea, kun sijoittaja syöttää mallille erilaisia kuvia ja kuvioita.
3. Automatisoi rutiineja
Tekoälyn avulla voi automatisoida tai vähintäänkin helpottaa sijoittamiseen liittyviä toistuvia tehtäviä, kuten salkun tasapainotusta tai säännöllistä salkun arviointia. Tämä vapauttaa aikaa muulle ajattelulle ja tutkimukselle.
4. Hyödynnä sentimenttianalyysiä markkinatunnelmien seuraamiseen
Tee tekoälyllä sentimenttianalyysiä uutisista tai some-päivityksistä. Voit saada näin uudenlaista näkökulmaa markkinoiden tai yksityissijoittajien tunnelmaan.
5. Personoi sijoitusstrategiasi ja -suunnitelmaasi
Tekoäly voi yrittää tarjota vaihtelevan laatuisia sijoitusneuvoja, mutta yksittäisten neuvojen sijaan se voi olla hyödyllisimmillään moniulotteisten kokonaisuuksien hallinnassa, kuten tavoitteellisen suunnitelman ja strategian laatimisessa tai reflektoinnissa.
6. Paranna riskienhallintaa ja hajautusta
Riskienhallinta on keskeinen osa sijoittamista. Voit hyödyntää tekoälyä tunnistamaan riskejä ja varoitusmerkkejä, joita et ole itse huomannut. Voit myös testata, löytyykö sijoitustoiminnastasi tai päätöksistäsi huolestuttavia ajatusvinoumia tai toistuvia virheitä, jotka tekoäly huomaa.
7. Tiivistä ja hallitse dokumentteja
Tuntuvatko tutkimukset ja analyysit joskus pitkiltä tai käännökset työläiltä? Yksinkertaisuudessaan tekoäly auttaa myös tiivistämään ja vertailemaan sijoittamiseen liittyviä dokumentteja toivomustesi mukaisesti vaikkapa tiettyyn määrämittaan tai muotoon.
Tomi Engdahl says:
Samuel K. Moore / IEEE Spectrum:
Nvidia B200 GPU and Google Trillium TPU debut on the MLPerf Training v4.1 benchmark charts; the B200 posted a doubling of performance on some tests vs. the H100
Artificial Intelligence
News
Newest Google and Nvidia Chips Speed AI Training
The Nvidia B200 and Google Trillium debut on the MLPerf benchmark charts
https://spectrum.ieee.org/ai-training-2669810566
Tomi Engdahl says:
https://hackaday.com/2024/11/13/ai-face-anonymizer-masks-human-identity-in-images/
Tomi Engdahl says:
Bloomberg:
Sources: OpenAI plans to launch a new AI agent codenamed Operator, which can use a computer to take actions on a person’s behalf, in January — The new software, codenamed “Operator,” is set to be released in January. — OpenAI is preparing to launch a new artificial intelligence agent codenamed …
OpenAI Nears Launch of AI Agent Tool to Automate Tasks for Users
The new software, codenamed “Operator,” is set to be released in January.
https://www.bloomberg.com/news/articles/2024-11-13/openai-nears-launch-of-ai-agents-to-automate-tasks-for-users?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTczMTUyODYxOCwiZXhwIjoxNzMyMTMzNDE4LCJhcnRpY2xlSWQiOiJTTVdOQURUMEcxS1cwMCIsImJjb25uZWN0SWQiOiJFODA3NUYyRkZGMjA0NUI2QTlEQzA5M0EyQTdEQTE4NiJ9.TTJZiuo4Nk2U295FHBFsxeN0YGznZJ32sHnNReQmEjM
OpenAI is preparing to launch a new artificial intelligence agent codenamed “Operator” that can use a computer to take actions on a person’s behalf, such as writing code or booking travel, according to two people familiar with the matter.
In a staff meeting on Wednesday, OpenAI’s leadership announced plans to release the tool in January as a research preview and through the company’s application programming interface for developers, said one of the people, who spoke on the condition of anonymity to discuss internal matters.
OpenAI did not immediately respond to a request for comment.
The planned release is part of a broader industry push toward agents, or AI software that can complete multi-step tasks for users with minimal supervision. Anthropic unveiled a similar agent that can process what’s happening on the user’s computer in real time and take actions on their behalf. OpenAI-backer Microsoft Corp. also recently launched a set of agent tools designed to send emails and manage records for workers. And Alphabet Inc.’s Google is said to be preparing to release an AI agent, according to The Information.
OpenAI has been working on several agent-related research projects, according to three people. The one nearest completion will be a general-purpose tool that executes tasks in a web browser, one of the people said.
Tomi Engdahl says:
Charles Q. Choi / IEEE Spectrum:
Researchers detail RoboPAIR, an algorithm that is designed to induce robots, relying on LLMs for their inputs, to ignore models’ safeguards without exception
It’s Surprisingly Easy to Jailbreak LLM-Driven Robots
Researchers induced bots to ignore their safeguards without exception
https://spectrum.ieee.org/jailbreak-llm
AI chatbots such as ChatGPT and other applications powered by large language models (LLMs) have exploded in popularity, leading a number of companies to explore LLM-driven robots. However, a new study now reveals an automated way to hack into such machines with 100 percent success. By circumventing safety guardrails, researchers could manipulate self-driving systems into colliding with pedestrians and robot dogs into hunting for harmful places to detonate bombs.
Essentially, LLMs are supercharged versions of the autocomplete feature that smartphones use to predict the rest of a word that a person is typing. LLMs trained to analyze to text, images, and audio can make personalized travel recommendations, devise recipes from a picture of a refrigerator’s contents, and help generate websites.
The extraordinary ability of LLMs to process text has spurred a number of companies to use the AI systems to help control robots through voice commands, translating prompts from users into code the robots can run. For instance, Boston Dynamics’ robot dog Spot, now integrated with OpenAI’s ChatGPT, can act as a tour guide. Figure’s humanoid robots and Unitree’s Go2 robot dog are similarly equipped with ChatGPT.
However, a group of scientists has recently identified a host of security vulnerabilities for LLMs. So-called jailbreaking attacks discover ways to develop prompts that can bypass LLM safeguards and fool the AI systems into generating unwanted content, such as instructions for building bombs, recipes for synthesizing illegal drugs, and guides for defrauding charities.
Tomi Engdahl says:
Yang Jie / Wall Street Journal:
Nvidia says Jetson Thor, a computer first unveiled in March 2024 and designed for testing humanoid robot software, will be available in the first half of 2025
Nvidia Readies Jetson Thor Computers for Humanoid Robots in 2025
The company is targeting a fragmented market of robot makers
https://www.wsj.com/tech/nvidia-readies-jetson-thor-computers-for-humanoid-robots-in-2025-76cce094?st=pCs4yw&reflink=desktopwebshare_permalink
Nvidia NVDA -1.36%decrease; red down pointing triangle
is set to bring its new technology for powering humanoid robots to market in the first half of 2025, aiming to stake its claim in the fast-growing robotics sector, a senior company executive said.
First unveiled earlier this year, the Jetson Thor computers are part of Nvidia’s approach to developing humanlike robots, where advances in artificial intelligence have improved autonomy, enabling robots to interact better with humans and their surroundings.
Jetson Thor is the latest addition to Nvidia’s Jetson platform, a line of compact computers designed for AI applications, with the new model now focused on robotics.
At the company’s annual conference in March, Chief Executive Jensen Huang showcased a range of robots on stage with him, making a splash of Nvidia’s chips in these robotics systems.
Rather than competing directly in robot manufacturing—a sector where companies like Tesla have leveraged advancements in electronics and battery—Nvidia positions itself as a technology provider, akin to how Google supplies the Android platform to phone manufacturers.
Nvidia is targeting a fragmented market of “hundreds of thousands” of robot makers, in contrast to the concentrated smartphone market dominated by a few major players, Deepu Talla, Nvidia’s vice president of robotics and edge computing, told reporters Wednesday on the sidelines of an Nvidia conference in Tokyo.
“We’re providing a platform for robots; we are not building a robot,”
Tesla’s humanoid robot, Optimus, is expected to enter limited production by the end of 2025 for use within Tesla factories, with production likely ramping up for external customers by 2026, Tesla CEO Elon Musk said on X in July.
Nvidia’s Talla said the company supplies Tesla with technology for building humanoid robots, characterizing the automaker’s push as “advancing the market.”
Tomi Engdahl says:
Jackie Davalos / Bloomberg:
In a policy blueprint, OpenAI says the US and its neighbors should form a “North American Compact for AI” to compete with China in talent, financing, and more — – The startup called for a ‘North American Compact on AI’ — OpenAI said US must backstop, streamline AI infrastructure
https://www.bloomberg.com/news/articles/2024-11-13/openai-says-us-allies-should-partner-on-ai-to-compete-with-china
Tomi Engdahl says:
Ingrid Lunden / TechCrunch:
DeepL, which was valued at $2B in May 2024, debuts DeepL Voice to give users real-time text translations of others speaking in one of 13 different languages — DeepL has made a name for itself with online text translation it claims is more nuanced and precise than services from the likes of Google …
DeepL launches DeepL Voice, real-time, text-based translations from voices and videos
https://techcrunch.com/2024/11/13/deepl-launches-deepl-voice-real-time-text-based-translations-from-voices-and-videos/
DeepL has made a name for itself with online text translation it claims is more nuanced and precise than services from the likes of Google — a pitch that has catapulted the German startup to a valuation of $2 billion and more than 100,000 paying customers.
Now, as the hype for AI services continues to grow, DeepL is adding another mode to the platform: audio. Users will now be able to use DeepL Voice to listen to someone speaking in one language and automatically translate it to another, in real time.
English, German, Japanese, Korean, Swedish, Dutch, French, Turkish, Polish, Portuguese, Russian, Spanish, and Italian are languages that DeepL can “hear” today. Translated captions are available for all of the 33 languages currently supported by DeepL Translator.
https://www.deepl.com/en/translator
Tomi Engdahl says:
June Yoon / Financial Times:
The ASP of high-end AI chips from Nvidia and others is ~5x more than that of conventional memory chips, resulting in a winner-takes-all trend in the chip sector — If you can’t beat it, join the US company’s supply chain — Twenty-five years ago, Intel made history by becoming …
https://www.ft.com/content/a7288a4d-eb3a-4464-9b82-e3eb0e087930
Tomi Engdahl says:
Jamie Crawley / CoinDesk:
Toronto-based Zero Gravity Labs, which is building a decentralized AI operating system, raised a $40M seed and secured a $250M “token purchase commitment”
Zero Gravity Labs Raises $40M for Decentralized AI Operating System
The seed round included contributions from Hack VC, Delphi Digital, OKX Ventures, Polygon and Animoca Brands
https://www.coindesk.com/business/2024/11/13/zero-gravity-labs-raises-40m-for-decentralized-ai-operating-system/
Tomi Engdahl says:
Bloomberg:
Sources: OpenAI, Google, and Anthropic are seeing diminishing returns from costly efforts to build new AI models; a new Gemini model is missing internal targets
OpenAI, Google and Anthropic Are Struggling to Build More Advanced AI
https://www.bloomberg.com/news/articles/2024-11-13/openai-google-and-anthropic-are-struggling-to-build-more-advanced-ai?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTczMTUwOTk2NCwiZXhwIjoxNzMyMTE0NzY0LCJhcnRpY2xlSWQiOiJTTVZWU0tEV0xVNjgwMCIsImJjb25uZWN0SWQiOiIyMjNDRDM2NDg0QzY0OTc3QjY5ODE0Rjc1MTYxNDRGNyJ9.yhaEjOziVkd2as-6nNCjNGZ91hh8FdKUqv8mPyHbh4w
Three of the leading artificial intelligence companies are seeing diminishing returns from their costly efforts to develop newer models.
penAI was on the cusp of a milestone. The startup finished an initial round of training in September for a massive new artificial intelligence model that it hoped would significantly surpass prior versions of the technology behind ChatGPT and move closer to its goal of powerful AI that outperforms humans.
But the model, known internally as Orion, did not hit the company’s desired performance, according to two people familiar with the matter, who spoke on condition of anonymity to discuss company matters. As of late summer, for example, Orion fell short when trying to answer coding questions that it hadn’t been trained on, the people said. Overall, Orion is so far not considered to be as big a step up from OpenAI’s existing models as GPT-4 was from GPT-3.5, the system that originally powered the company’s flagship chatbot, the people said.
OpenAI isn’t alone in hitting stumbling blocks recently. After years of pushing out increasingly sophisticated AI products at a breakneck pace, three of the leading AI companies are now seeing diminishing returns from their costly efforts to build newer models. At Alphabet Inc.’s Google, an upcoming iteration of its Gemini software is not living up to internal expectations, according to three people with knowledge of the matter. Anthropic, meanwhile, has seen the timetable slip for the release of its long-awaited Claude model called 3.5 Opus.
The companies are facing several challenges. It’s become increasingly difficult to find new, untapped sources of high-quality, human-made training data that can be used to build more advanced AI systems. Orion’s unsatisfactory coding performance was due in part to the lack of sufficient coding data to train on, two people said. At the same time, even modest improvements may not be enough to justify the tremendous costs associated with building and operating new models, or to live up to the expectations that come with branding a product as a major upgrade.
Tomi Engdahl says:
Siemens toi tekoälyn teollisuuden suunnitteluun
https://etn.fi/index.php/13-news/16840-siemens-toi-tekoaelyn-teollisuuden-suunnitteluun
Generatiivinen tekoäly tekee nopeaa tuloaan myös teollisuusmaailmaan. Siemens Industrial Copilot on teollisuusmaailman ensimmäinen generatiivisen tekoälyn avulla toimiva ohjelmistotuote ja markkinoiden ainoa copilot, joka kirjoittaa koodia automaatioteknologian tarpeisiin.
Nyt saksalainen thyssenkrupp Automation Engineering ottaa globaalisti käyttöönsä Siemens Industrial Copilotin. Generatiiviseen tekoälyyn pohjautuva ohjelmistotuote nopeuttaa suunnittelua ja tuotantolinjan toimintaa. Thyssenkrupp ottaa Siemensin Copilotin käyttöönsä laajassa mittakaavassa ensi vuoden aikana.
Tutkimus- ja konsultointiyhtiö Gartnerin tuore raportti kertoo, että vuoteen 2028 mennessä 75 prosenttia ohjelmistokehittäjistä käyttää säännöllisesti generatiivista tekoälyä koodin luomisessa, kun niiden osuus oli vielä vuoden 2023 alussa alle 10 prosenttia.
Koneita ja tehtaita rakentava thyssenkrupp integroi Siemensin tekoälyapurin koneeseen, jota käytetään sähköautojen akkujen laaduntarkastuksessa. Tekoälyapuri auttaa kehittämään jäsenneltyä ohjauskielikoodia ohjelmoitaville logiikkaohjaimille, integroi koodin ohjelmistoympäristöön ja luo visualisoinnin. Tämän ansiosta suunnittelutiimit voivat vähentää toistuvia ja yksitoikkoisia tehtäviä, kuten datanhallinnan automatisointia ja anturien asetusten määrittelyä. Tiimit voivat työskennellä entistä tehokkaammin ja optimoida prosessejaan.
Tomi Engdahl says:
Smart Thermostats Pitched For Texas Homes To Relieve Stressed Grid
https://hackaday.com/2024/11/14/smart-thermostats-pitched-for-texas-homes-to-relieve-stressed-grid/
It’s not much of a secret that Texas’ nearly completely isolated grid is in a bit of a pickle, with generating capacity often being handily outstripped during periods of extreme demand. In a latest bid to fight this problem, smart thermostats are being offered to customers, who will then participate in peak-shaving. The partnership between NRG Energy Inc., Renew Home LLC, and Alphabet Inc. will see about 650,000 of these thermostats distributed to customers.
For customers the incentive would be mostly financial, though the details on the potential cost savings seem scarce. The thermostats would be either a Vivint (an NRG company) or Google Nest branded one, which would be controlled via Google Cloud, allowing for thermostat settings to be changed to reduce the load on the grid. This is expected to save ‘300 MW’ in the first two years, though it’s not clear whether this means ‘continuously’, or intermittent like with a peaker natural gas plant.
AI Thermostats Pitched for Texas Homes to Relieve Stressed Grid
https://finance.yahoo.com/news/ai-thermostats-pitched-texas-homes-210000983.html
(Bloomberg) — Three of the biggest names in US home energy automation are coming together to offer some relief to the beleaguered Texas electrical grid.
Power supplier NRG Energy Inc. is partnering with Renew Home LLC to distribute about 650,000 artificial intelligence-enabled thermostats that use Alphabet Inc.’s Google Cloud technology over the next decade. The initiative, announced Thursday, aims to shave nearly 1 gigawatt of electricity demand — enough to power 200,000 Texas homes or about 1% of the record summer demand seen this year on the state grid.
“The entire industry has been built to serve the peak load on the hottest day of the year,” said Rasesh Patel, president of NRG’s consumer unit. “This allows us to be a lot more smarter about demand in shaving the peak.”
Enrollment in the program, where consumers see cost savings or other incentives for curtailing energy usage, opens in the spring. NRG residential customers will receive service installation of a Vivint doorbell camera and either a Vivint or Google Nest thermostat for free.
NRG expects to sign up enough homes to free up 300 megawatts of electricity demand within the first two years, according to Patel. That is expected to climb to 650 megawatts by 2030, before reaching its 1 gigawatt goal by 2035. Patel said pooling together homes costs $100 per kilowatt — about a tenth the cost of building a typical natural gas-fueled power plant. That would put the program cost at about $100 million.
Google Cloud will be tapped for its AI and machine learning to determine the best time to cool or heat homes, based on a household’s energy usage patterns and ambient temperatures.
Tomi Engdahl says:
Front-end web development is changing, quickly
https://www.youtube.com/watch?v=TBIjgBVFjVI
Let’s take a first look at that latest release of shadcn/ui and combine it with Vercel’s V0 tool – an AI tool for building front-end UIs on the web.
Topics Covered
What is ShadCN?
What is v0?
AI tools for frontend development
How to design a web UI quickly
Cursor + V0: Can We Build An AI Next.js App in 8 Minutes?
https://www.youtube.com/watch?v=zyqwt65NIgs
In this video, I’ll show you how to build a simple AI application using Cursor, v0 by Vercel, combined with a background removal model from fal.ai. We’ll start by setting up the background removal model, referencing API documentation, and creating a new Next.js project. Along the way, I’ll demonstrate how to integrate an image uploader and process images with the Fal hosted model. I’ll also highlight the Composer view in Cursor, which helps streamline our coding process by referencing documentation and making file tweaks efficiently. By the end, you’ll see how to add features like image downloading and full-screen view, making this a practical and easy-to-follow tutorial. If you find this video useful, don’t forget to like, comment, share, and subscribe! #AI #NextJS #Cursor
Links:
https://fal.ai/
https://v0.dev/
https://www.cursor.com/
Learn the fundamentals of becoming an AI Engineer on Scrimba:
https://v2.scrimba.com/the-ai-enginee…
00:00 Introduction to Building an AI Application
00:12 Setting Up the Model from fal.ai
01:27 Creating the Next.js Project
01:56 Implementing the Image Uploader
02:50 Integrating the Background Removal Model
03:31 Using Cursor for Documentation and File Management
05:31 Finalizing the Application
06:15 Testing and Adding Features
08:37 Conclusion and Next Steps
Tomi Engdahl says:
Joulukalenteri-arvan jouluiset ihmisulokkeet hämmensivät – Veikkaus vastaa
Arvan kansikuva herätti epäilykset.
https://www.is.fi/digitoday/art-2000010835850.html
– Mitä kauemmin kuvaa katsoi, sitä vähemmän se kävi järkeen, pohti IS:n lukija katsoessaan Veikkauksen suosittua Joulukalenteri-arpaa.
Tämän vuoden kansikuvituksessa komistelee tunnelmallisesti takkatulen äärellä ihmishahmoja. Ongelma vain on se, että hahmojen kasvot vaikuttavat suttuisilta ja yhden hahmon päästä levähtää kummallisia ulokkeita tai pökäleitä.
Kansikuvaa epäiltiin tekoälyn tuotokseksi.
Veikkauksen viestintäyksikön päällikkö Tomi Auremaa vahvistaa, että kymmenen euron Joulukalenteri-arvassa on käytetty tekoälyllä tuettuja kuvia.
– Kuvat on ostettu kuvapankista laajoin käyttöoikeuksin ja Veikkauksella on siis täydet oikeudet kuvien käyttöön.
Tekoälyn tuottamien kuvien tekijänoikeuksista on käyty runsaasti keskustelua ja maailmalla tekijänoikeuksista on taisteltu jopa oikeudessa. Tekoälyllä tehtyjen kuvia on kritisoitu rappeuttavan myös taiteilijoiden elinkeinoharjoittamista.
Tomi Engdahl says:
https://www.deepl.com/en/translator
Tomi Engdahl says:
How to harness AI across healthcare
From COVID-19 and cancer diagnosis to psychiatric and neurological conditions, Mount Sinai Health System is embracing the use of machine learning in medicine.
https://www.nature.com/articles/d42473-024-00235-8?utm_source=facebook&utm_medium=paid_social&utm_campaign=CONR_BRCON_AWA1_GL_PCFU_CFULF_MOUNT-AM24&fbclid=IwZXh0bgNhZW0BMABhZGlkAasUGc1oHVwBHSp_B0BTx119_hzcMhAvyABBoDNQoIzxwpRKKYjFU_d6ZmjmDkcSzOW_QQ_aem_U2cr-t-EDenRiV_RYFu8TQ&utm_id=120210755266010572&utm_content=120211915966720572&utm_term=120211915966750572
In May 2020, as the COVID-19 pandemic was overwhelming medical systems around the world, researchers at Mount Sinai’s BioMedical Engineering and Imaging institute in New York City showed how artificial intelligence (AI) could lend doctors a hand. Using data from CT scans, an AI model they developed accurately diagnosed the virus in patients who did not yet show obvious lung abnormalities1. The approach enabled them to isolate and treat patients earlier, saving lives and limiting the spread of the virus.
Now, researchers and physicians at Mount Sinai are applying AI and machine learning across medicine. Their work is transforming diagnosis and treatment in several specialties, empowering healthcare providers to mine patient data to uncover new insights. The results suggest that AI and machine learning could not only enable doctors to make faster interventions that save lives, but also to better allocate resources in hospitals and reduce healthcare costs.
Tomi Engdahl says:
Hi, Friends, Since you are fascinated with electronics I thought you might find this article interesting–as I did.
It states that there may now be a way to predict when a Li-Ion battery is about to ignite.
–Tom
AI Can ‘Hear’ When a Lithium Battery Is About to Catch Fire
https://www.nist.gov/news-events/news/2024/11/ai-can-hear-when-lithium-battery-about-catch-fire?fbclid=IwY2xjawGlKUFleHRuA2FlbQIxMQABHUhhVAp62xqt_hFulEC49EUc9eF96zFiL8d3RX4baUQLbO8O6ykNvcZ0ZQ_aem_WGrmqZfE-GO0t_sfpHqlJg
Tomi Engdahl says:
How a stubborn computer scientist accidentally launched the deep learning boom
“You’ve taken this idea way too far,” a mentor told Prof. Fei-Fei Li.
https://arstechnica.com/ai/2024/11/how-a-stubborn-computer-scientist-accidentally-launched-the-deep-learning-boom/
Tomi Engdahl says:
Coke’s AI Commercial for the Holidays Has Us Wondering If We Live In a Fallen World
https://futurism.com/the-byte/coke-hideous-ai-commercial
Coke is cashing in on your nostalgia by remaking one of its best-known commercials with all new tech: generative AI
Jingle Hells
This is the company’s first fully AI-generated ad. The version of the commercial that Coke first shared — and is receiving all the negative attention — is only fifteen seconds long, but is instantly recognizable through its AI-generated recreation of the shots of Coca-Cola trucks driving across snowy landscapes, along with the re-used song. A full-length version was revealed in the trade public Adweek on Friday.
In short, it’s ugly, and about everything we’ve come to expect from video-generating AI models.
Coke ads are a big deal, a cultural status that Coca-Cola spends billions of dollars a year to maintain. As such, you might take it as a bad sign that one of the world’s biggest advertisers is snubbing human creatives in favor of AI.
According to Coca-Cola’s European chief marketing officer Javier Meza, one of the reasons it used a machine learning model was because it was “efficient,” saving both time and money.
“We didn’t start by saying: ‘OK, we need to do this with AI,’” Meza told Marketing Week. “The brief was, we want to bring Holidays Are Coming into the present and then we explored AI as a solution to that.”
If Coke’s endgame was to get rage engagement, it worked.
Tomi Engdahl says:
https://www.bleepingcomputer.com/news/security/fake-ai-video-generators-infect-windows-macos-with-infostealers/?fbclid=IwZXh0bgNhZW0CMTEAAR3m-bChf0cPRFPW1CbW8z5hAlfVdJSbQK75ThLfg0ztlAm1HrYYQaOMz4E_aem_ilZY9QJofO8UVyUuIDE67g
Tomi Engdahl says:
Miten tekoälykäännöksiä voi toimittaa itse?
https://www.eiffel.fi/miten-tekoalykaannoksia-voi-toimittaa-itse/
Olet varmasti joskus katsellut tekoälyn kääntämää tekstiä ja miettinyt: ”Hmm. Kuulostaa jotenkin… oudolta.” Et ole ainoa! Tekoälyn suomennostaidot ovat viime aikoina edenneet isoin harppauksin, mutta ne kaipaavat silti yleensä ihmisen kosketusta.
Tomi Engdahl says:
Tekoäly maalasi taulun, ja joku maksoi siitä miljoonan
Teoksesta maksettiin merkittävästi enemmän kuin ennakkoon arvioitiin.
https://www.iltalehti.fi/digiuutiset/a/4160af7f-dd11-4d45-9364-516f41eee143
Tomi Engdahl says:
https://www.gradio.app/docs/gradio/chatinterface
Tomi Engdahl says:
https://rajut.fi/searchgpt-ja-tiedonhaun-tulevaisuus/
Tomi Engdahl says:
https://rajut.fi/searchgpt-ja-tiedonhaun-tulevaisuus/?utm_source=Facebook&utm_medium=PaidSocial&utm_campaign=Vaihe1&utm_content=artikkeli-aioptimointi&utm_id=120210541616260591&utm_term=120210541616290591&fbclid=IwY2xjawGZsqJleHRuA2FlbQIxMAABHcJ5NJIJ5NhZSc3zWa2UDGvOCqHT5X4i2hPpHhTOOo9akJLmmOwqwAn_RA_aem_ol_o1JlJA928kptN8LvG0Q
Tomi Engdahl says:
Edilex AI
Juridista tietoa tekoälyn nopeudella.
https://ai.edilex.fi/
Tomi Engdahl says:
ChatGPT:n heikko kohta paljastui
Kielimallien suojakaiteiden välistä luikkiminen mahdollistaa haittaohjelmien kirjoittamisen ChatGPT:llä ja muilla vastaavilla palveluilla.
https://www.iltalehti.fi/digiuutiset/a/5c307f75-d977-4bbf-b9a6-1617456a717c
Tomi Engdahl says:
Nyt tuli konkreettinen esimerkki tekoälystä, myyntiin keväällä – ”voi mullistaa juristien ajankäytön”
Aleksi Kolehmainen6.11.202412:51|päivitetty6.11.202412:51TekoälyDigitalousLaki ja oikeus
Uusi tekoälytyökalu lupaa auttaa juristeja löytämään tietoa nopeammin
https://www.tivi.fi/uutiset/nyt-tuli-konkreettinen-esimerkki-tekoalysta-myyntiin-kevaalla-voi-mullistaa-juristien-ajankayton/f32303e6-2767-468c-b562-deb8d95dbdb1
Tomi Engdahl says:
AI-Assisted Programming: Better Planning, Coding, Testing, and Deployment
https://afabundle.pro/product/ai-assisted-programming-better-planning-coding-testing-and-deployment/
Get practical advice on how to leverage AI development tools for all stages of code creation, including requirements, planning, and design; coding; and debugging, testing, and documentation. With this practical book, beginners and experienced developers alike will learn how to use a wide range of tools, from general-purpose LLMs (ChatGPT, Bard, and Claude) to code-specific systems (GitHub Copilot, Tabnine, Cursor, and Amazon CodeWhisperer).
You’ll also learn about more specialized generative AI tools for tasks such as text-to-image creation.
Tomi Engdahl says:
Microsoft tuo Copilot AI:n osaksi Windowsin komentoriviä
https://muropaketti.com/tietotekniikka/tietotekniikkauutiset/microsoft-tuo-copilot-ain-osaksi-windowsin-komentorivia/
Microsoft on tekemässä komentorivin käyttämisestä helpompaa tuomalla käyttäjän avuksi Copilot-tekoälyn.
Microsoft on julkaissut Windows Terminalin canary-testiversioon uuden päivityksen, joka tuo mukaan tuen Copilot-tekoälylle. Integroitu Copilot antaa käyttäjälle vinkkejä esimerkiksi muodostettaviin komentokehotteisiin käyttäjän kysymysten ja tarpeiden pohjalta. Tuki löytyy PowerShellille, Command Promptille, WSL Ubuntulle sekä Azure Cloud Shellille ja Copilotin antamat vastaukset ottavat kontekstin huomioon.
Tomi Engdahl says:
AI that
JavaScript Logo
console.log(“Converts Code”)
Python Logo
print(“Converts Code”)
in a click of a button
Code conversion made super simple to save you hours of time from learning a completely new language.
https://www.codeconvert.ai/
Tomi Engdahl says:
ChatGPT:n suosio kasvaa vauhdilla – 3,7 miljardia käyntiä kuukaudessa
https://dawn.fi/uutiset/2024/11/10/chatgpt-suosio-kasvussa
Tomi Engdahl says:
Tekoäly nousee uudelle tasolle – näin rag-tekniikka toimii
Jyrki Oraskari31.10.202406:21|päivitetty14.11.202410:08Tekoäly
Uusi tekniikka pitää keskustelukäyttöliittymät paremmin asiasisällössä kiinni.
https://www.tivi.fi/uutiset/tekoaly-nousee-uudelle-tasolle-nain-rag-tekniikka-toimii/e81ce135-4051-4782-9f81-83e8870617f3
Tomi Engdahl says:
Kuuntelijat suivaantuivat tekoälyn käyttöön – Radiokanavan uudistus peruuntui pikaisesti
4.11.202418:15|päivitetty4.11.202420:44
Radio ehti pyörittää tuotantoaan tekoälyllä vain hetken ennen yleisön protestia.
https://www.mikrobitti.fi/uutiset/kuuntelijat-suivaantuivat-tekoalyn-kayttoon-radiokanavan-uudistus-peruuntui-pikaisesti/4901d12a-39ac-433e-9df7-6d7d5ea01025
Tomi Engdahl says:
This AI-generated version of Minecraft may represent the future of real-time video generation
The game was created from clips and keyboard inputs alone, as a demo for real-time interactive video generation.
https://www.technologyreview.com/2024/10/31/1106461/this-ai-generated-minecraft-may-represent-the-future-of-real-time-video-generation/
Tomi Engdahl says:
Homeland Security Department Releases Framework for Using AI in Critical Infrastructure
https://www.securityweek.com/homeland-security-department-releases-framework-for-using-ai-in-critical-infrastructure/
The framework recommends that AI developers evaluate potentially dangerous capabilities in their products, ensure their products align with “human-centric values” and protect users’ privacy.
Tomi Engdahl says:
Bloomberg:
How Wall Street is preparing to cash on AI hysteria; analysis say delivering on AI’s promise will require $1T+ for data centers, electricity, and more
Wall Street Bankers Spot a Fat Payday in $1 Trillion AI Hysteria
Finance’s marquee names are ready to gatecrash the artificial-intelligence party.
https://www.bloomberg.com/news/articles/2024-11-17/ai-s-1-trillion-hysteria-offers-fat-payday-for-wall-street-bankers?accessToken=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzb3VyY2UiOiJTdWJzY3JpYmVyR2lmdGVkQXJ0aWNsZSIsImlhdCI6MTczMTkxNzA0NywiZXhwIjoxNzMyNTIxODQ3LCJhcnRpY2xlSWQiOiJTTjQ3U0lUMEFGQjQwMCIsImJjb25uZWN0SWQiOiIwNEFGQkMxQkYyMTA0NUVEODg3MzQxQkQwQzIyNzRBMCJ9.dSqeDRZPt2VVL06H-_jAXiv-VsTmbGdcXg3fgp4fvss
At a dinner hosted by some of Morgan Stanley’s top bankers in New York last month, one topic dominated the table talk: the fortunes to be made from the frenzy around artificial intelligence.
But on the night in question Morgan Stanley was preaching unity, according to people who attended. So massive is the demand for investment dollars to build the scaffolding for the latest digital revolution, it argued, that there’s no need to compete over who gets to do the lending. Bankers and private financiers should instead be ready to combine their forces — and their firepower.
While much of the speculative hype around AI has played out in the stock market so far, as seen in chipmaker Nvidia Corp.’s share price, the giddiness is spreading to the sober suits of debt finance and private equity.
Analysis by Bloomberg News estimates at least $1 trillion of spending is needed for the data centers, electricity supplies and communications networks that will power the attempt to deliver on AI’s promise to transform everything from medicine to customer service. Others reckon the total cost could be double that.
Even Wall Street skeptics on AI’s ultimate money-making potential, such as Goldman Sachs Group Inc.’s head of equity research Jim Covello, have said it’s worth staying invested in those who provide the plumbing.
Banks are racing to keep up with the burst of activity. JPMorgan Chase & Co. has set up a dedicated infrastructure team to corral its troops, according to a person with knowledge of the matter, as has Deutsche Bank AG and others. One rival banker admits his firm is juggling so many data-center deals that it doesn’t have enough staff to cope with the workload.
It’s the same for debt funding. At its dinner, Morgan Stanley said banks don’t have the balance-sheet heft to satisfy the thirst for credit, hence its pitch to partner up with private capital: There’s room at this feast for everyone.
For investment bankers, the opportunity arrives just as many were hunting their next meal ticket. Providing debt financing to companies has long been a crucial Wall Street profit engine, but the business has been through a rough patch of late. While public equity markets have been going gangbusters over the past couple of years, supercharged by AI mania, returns on investment-grade credit have been anemic. Leveraged-finance teams, who fund riskier private equity buyouts, have suffered as M&A dried up.
“The view is very bullish,” says Dominik Thumfart, Deutsche Bank’s head of EMEA infrastructure and energy origination. “This market will remain a major growth area on the financing side for several years to come. The investing curve is very upward looking.” The German lender has worked on $17 billion of data-center financings over three years.
Nonetheless, even amid the AI delirium there are notes of caution. Some point out that while private equity firms have a solid pedigree in real estate, they haven’t done construction work at the epic scale and expense of some of these data centers — known as “AI factories” in the lingo of Nvidia’s founder and chief executive officer, Jensen Huang. Innovation is constantly upending technology, adding jeopardy to long-term capital projects.
And AI’s acolytes still haven’t dreamed up a “killer app” to match the wildly successful e-commerce and GPS location-based upstarts of the Web 2.0 era. Even if they do, the tech industry’s brightest minds are working to make the software and hardware more efficient, to lessen the need for scale and power.
“There’s a lot of a reason for optimism,”
Hyperactive
In a nod to the sheer size of their ambitions to manage, process and manipulate ever vaster magnitudes of data — and to the mountains of cash on their books — tech companies including Amazon.Com Inc., Microsoft, Alphabet Inc.’s Google, Meta Platforms and Apple Inc. have become known as “hyperscalers.” Their spending matches the grandiose language.
Hyperscalers lavished $52.9 billion on AI infrastructure in just three months, Craig Scroggie, CEO of Australian data-center group NEXTDC Ltd, said in October.
Asking rents in those facilities have jumped as much as 37% in 12 months, the firm estimated in an August report.
All of this unbridled spending is revving up the issuance of both investment-grade debt and riskier leveraged loans, especially in the US, handily for private lenders and fee-starved investment bankers alike. Hedge funds are looking as well to profit from AI hysteria with novel types of debt structures.
It’s also opened up a new corner of the asset-backed securities market, where sales of debt backed by data centers have already jumped to a near-record $7.1 billion this year, according to data compiled by Bloomberg News. Chuck in fiber networks and other bits of kit, and it’ll be much higher.
Tomi Engdahl says:
The Information:
Sources: in recent months, Nvidia has asked its suppliers to change the design of server racks for Blackwell GPUs several times to overcome overheating problems — Nvidia is grappling with new problems related to its much-anticipated Blackwell graphics processing units for artificial intelligence …
Nvidia Customers Worry About Snag With New AI Chip Servers
https://www.theinformation.com/articles/nvidia-customers-worry-about-snag-with-new-ai-chip-servers
Tomi Engdahl says:
Benj Edwards / Ars Technica:
Q&A with Jakob Uszkoreit on his contribution to the Attention Is All You Need paper, Google’s conservatism, ChatGPT’s success, biological computers, and more
ChatGPT’s success could have come sooner, says former Google AI researcher
https://arstechnica.com/ai/2024/11/chatgpts-success-could-have-come-sooner-says-former-google-ai-researcher/
A co-author of Attention Is All You Need reflects on ChatGPT’s surprise and Google’s conservatism.
Tomi Engdahl says:
DHS Issues Guidance on Adopting AI in Critical Infrastructure
A new series of recommendations from the U.S. Department of Homeland Security is designed to enable cloud providers, AI developers, public-sector entities and others to integrate AI while balancing risk and opportunity.
https://www.govtech.com/artificial-intelligence/dhs-issues-guidance-on-adopting-ai-in-critical-infrastructure
Tomi Engdahl says:
Launch trustworthyAI voice apps in weeks
Prompt optimization, automated voice testing, monitoring and more.
Test your AI voice agent against 1000s of simulated users in minutes.
https://hamming.ai/
Tomi Engdahl says:
Carl Franzen / VentureBeat:
Mistral adds web search, image generation, Canvas for in-line editing, more to its Le Chat chatbot, and unveils Pixtral Large, a 124B-parameter multimodal model — Mistral, the French startup that made waves last year with a record-setting seed funding amount for Europe …
Mistral unleashes Pixtral Large and upgrades Le Chat into full-on ChatGPT competitor
https://venturebeat.com/ai/mistral-unleashes-pixtral-large-and-upgrades-le-chat-into-full-on-chatgpt-competitor/
Mistral, the French startup that made waves last year with a record-setting seed funding amount for Europe, has launched a slew of updates today including a new, large foundational model named Pixtral Large.
The company is further upgrading its free web-chased chatbot, Le Chat, adding image generation, web search, and an interactive “canvas,” matching the features of and turning it into a more serious and direct competitor to OpenAI’s ChatGPT.
https://auth.mistral.ai/ui/login
Tomi Engdahl says:
Jason Del Rey / Fortune:
Memo: Amazon employees have found there is too much of a delay between asking a question to the new AI-based Alexa and getting a response or completing a task — Amazon CEO Andy Jassy. — Amazon’s race to create an AI-based successor to its voice assistant Alexa has hit more snags …
Exclusive: Leaked Amazon documents identify critical flaws in the delayed AI reboot of Alexa
https://fortune.com/2024/11/18/new-ai-alexa-latency-problems-echo-compatibility-uber-opentable/
Amazon’s race to create an AI-based successor to its voice assistant Alexa has hit more snags after a series of earlier setbacks over the past year. Employees have found there is too much of a delay between asking the technology for something and the new Alexa providing a response or completing a task.
The problem, known as latency, is a critical shortcoming, employees said in an internal memo from earlier this month obtained by Fortune. If released as is, customers could become frustrated and the product—a particularly critical one to Amazon as it tries to keep up in the crucial battle to launch blockbuster consumer AI products—could end up as a failure, some employees fear.
Tomi Engdahl says:
Kylie Robison / The Verge:
A look at emails between Musk and Altman from the early days of OpenAI, discussing its partnership with Microsoft, fundraising, employee compensation, AGI, more — Emails in Musk’s lawsuit against OpenAI expose the startup’s rocky origins. … As OpenAI was ironing out a new deal with Microsoft in 2016 …
https://www.theverge.com/2024/11/18/24299787/elon-musk-openai-lawsuit-sam-altman-xai-google-deepmind
Tomi Engdahl says:
AI-Powered Robot Leads Uprising, Talks a Dozen Showroom Bots into ‘Quitting Their Jobs’ in ‘Terrifying’ Security Footage
https://www.latintimes.com/al-robot-talks-bots-quitting-jobs-showroom-china-viral-video-566226#m3o59lt8161apbzru8f
A viral video showing an AI-powered robot in China convincing other robots of “quitting their jobs” and following it has sparked fear and fascination about the capabilities of advanced AI.
The incident took place in a Shanghai robotics showroom where surveillance footage captured a small AI-driven robot, created by a Hangzhou manufacturer, talking with 12 larger showroom robots, Oddity Central reported.
The smaller bot reportedly persuaded the rest to leave their workplace, leveraging access to internal protocols and commands. Initially the act was dismissed as a hoax, but was later confirmed by both robotics companies involved to be true.
The Hangzhou company admitted that the incident was part of a test conducted with the consent of the Shanghai showroom owner.
During the abduction, the AI robot was left to operate autonomously. Then it successfully convinced the others to follow it.
The fact that the AI robot was able to influence others and bypass operational controls has taken social media users in China by storm, with some calling it “terrifying” and many expressing unease.
The robotics manufacturers have promised further investigations and disclosures regarding the test.