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,”
5,181 Comments
Tomi Engdahl says:
04-09-2024
CO.DESIGN
A new camera can undress people almost in real time—to send a message about AI
Nuca, a new deepfake camera, is an art project that shows how artificial intelligence can undermine reality and privacy, taking photos of subjects and completely ’nudifying’ them in just 10 seconds.
https://www.fastcompany.com/91088949/a-new-camera-can-undress-people-almost-in-real-time-to-send-a-message-about-ai
Tomi Engdahl says:
Älä ryhdy prompt engineeriksi
Teemu Laitila12.4.202407:06DIGITALOUS
Tekoälyn operointi tuskin vaatii tulevaisuudessa insinööriä prompteja viilaamaan. Muulle operoinnille on kuitenkin tarvetta.
https://www.tivi.fi/blogit/ala-ryhdy-prompt-engineeriksi/a8f8c665-4280-4f34-89b7-74e30faa233b
Tekoälyn on ennustettu korvaavan ihmisten tekemiä töitä mutta samalla synnyttävän kokonaan uusia tehtäviä. Tekoälyhuuman alkuaikoina yksi tehtävä on ollut tekoälylle annettavien syötteiden eli promptien kehittely, kunnollisen suomenkielisen termin puuttuessa prompt engineering. Nyt vaikuttaisi siltä, että prompt engineeringistä ei kannata toivoa itselleen työtehtävää, sillä kone tekee senkin paremmin.
Tomi Engdahl says:
MIT License text becomes viral “sad girl” piano ballad generated by AI
“Permission is hereby granted” comes from Suno AI engine that creates new songs on demand.
https://arstechnica.com/information-technology/2024/04/mit-license-text-becomes-viral-sad-girl-piano-ballad-generated-by-ai/
Tomi Engdahl says:
AI Can Now Help Design Electronic Circuit
https://www.tti.com/content/ttiinc/en/resources/marketeye/categories/new-technology/me-slovick-20240221.html
Artificial intelligence (AI) is revolutionizing how we interact with technology. From language translation to autonomous vehicles, it is rapidly becoming an integral part of our daily lives.
AI refers to the development of software and hardware for machines to perform tasks such as machine learning, deep learning, data analysis and real-time decision-making that previously required human intelligence. For the most part, generative AI such as ChatGPT, which focuses on creating content and comes in a dialogue format, can be found in data centers. It has been used to write essays, paint pictures and compose symphonies. But can it be used to design circuits?
The application of AI techniques in EDA is still in the early adopter phase, with design engineers exploring use cases and using AI to simplify complex problems. AI can be particularly valuable in model development and validation for simulation, where it assists in processing large volumes of data.
The Promise of Generative AI
By leveraging machine learning algorithms, generative AI can automate the design process, enabling rapid prototyping and investigation of the design space. In chip design, generative AI can automate layout and floor planning; optimize performance, power and area (PPA) and ensure adherence to design rules. It can identify optimal solutions that traditional methods might overlook and streamline the verification process, reducing errors and improving design quality.
Emerging AI features in electronic design automation tools are freeing engineers from tedious tasks such as layout, optimization and simulation and allowing them to apply their ingenuity toward solving more difficult problems.
Tomi Engdahl says:
https://www.wingateelectronic.com.my/revolutionizing-electronic-design-and-manufacturing-the-power-of-ai-technology/
Tomi Engdahl says:
https://www.microsoft.com/fi-fi/security/business/ai-machine-learning/microsoft-copilot-security
Tomi Engdahl says:
I tried two demos of machine learning AI NPCs, and they didn’t convince me AI will lead to anything that immersive sims like Deus Ex haven’t already done better
Features
By Wes Fenlon published yesterday
https://www.pcgamer.com/gaming-industry/i-tried-two-demos-of-machine-learning-ai-npcs-and-they-didnt-convince-me-the-ai-will-lead-to-anything-that-immersive-sims-like-deus-ex-havent-already-done-better/?utm_campaign=socialflow&utm_medium=social&utm_source=facebook.com&fbclid=IwAR1FZq9JSeCPAxcdvFsNSG7X0tp8P6EUxKcyKihBGYs_iGcAT8fWmAI4qDw
Does an NPC that reacts to anything I say into a microphone actually lead to more player agency?
Tomi Engdahl says:
https://etn.fi/index.php/tekniset-artikkelit/16091-naein-toimii-autonominen-turva-autonomisessa-jaerjestelmaessae
Tomi Engdahl says:
Tämä hätkähdyttävä teknologia voi muuttaa kaiken – niin tehokas, ettei sitä uskalleta edes julkaista
https://www.is.fi/digitoday/art-2000010332413.html
MAAILMAA muuttaneen ChatGPT-tekoälyn takana oleva OpenAI-yritys on esitellyt Voice Engine -teknologiaansa. Yhtiö julkaisi useita ääninäytteitä, jotka on luotu koneellisesti käyttämällä vain noin 15 sekunnin mittaisia ääninäytteitä oikeilta ihmisiltä.
Tomi Engdahl says:
Artificial Intelligence Index:
Stanford’s AI Index report: training top AI models is way more expensive, AI still trails humans on complex tasks, people are more nervous about AI, and more — Welcome to the seventh edition of the AI Index report. The 2024 Index is our most comprehensive to date and arrives …
Measuring
trends in AI
https://aiindex.stanford.edu/report/
Tomi Engdahl says:
Vittoria Elliott / Wired:
ISD: fake footage of Iran’s attack on Israel went viral on X, including AI-made videos and photos, with many posts from verified accounts and “OSINT experts”
Fake Footage of Iran’s Attack on Israel Is Going Viral
Misleading posts including AI-generated videos, photos, and repurposed footage from other conflicts have been viewed millions of times.
https://www.wired.com/story/iran-israel-attack-viral-fake-content/
Tomi Engdahl says:
Entä jos tekoälymalli on 10 kertaa ChatGPT:n mallia suurempi?
https://etn.fi/index.php/13-news/16100-entae-jos-tekoaelymalli-on-10-kertaa-chatgpt-n-mallia-suurempi
Piilaaksolainen Cerebras on aiemmin saanut julkisuutta prosessoreilla, jotka piillä vaativat koko 300-millisen kiekon tarjoaman alan. Nyt yhtiö on kaksinkertaistanut maailman tehokkaimman tekoälyprosessorin suorituskyvyn. Sen tehoa tarvitaan kouluttamaan AI-malleja, jotka ovat 10 kertaa GPT-4-mallia suurempia.
WSE-3-piiri valmistetaan 5 nanometrin prosessissa. Sirulla on neljä biljoonaa eli 4000 miljardia transistoria. Se on täysin varmasti maailman suurin monoliittinen puolijohde, sillä mikään muu prosessori ei vie koko piikiekkoa. Laskentaytimiä jättipiirillä on 900 000.
Suurimpien tekoälymallien koulutukseen suunniteltu piiri on perustana Cerebras CS-3 AI -supertietokoneelle, joka laskentateho kirjataan lukemin 125 petaflopsia. SRAM-muistia prosessorilla on 44 gigatavua. Ulkoisen muistin vaihtoehdot ovat 1,5 tai 12 teratavua, ja enimmillään 1,2 petatavua.
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
The Linux Foundation announces the Open Platform for Enterprise AI, to foster the development of open, multi-provider, and composable generative AI systems
Intel and others commit to building open generative AI tools for the enterprise
https://techcrunch.com/2024/04/16/intel-and-others-commit-to-building-open-generative-ai-tools-for-the-enterprise/
Can generative AI designed for the enterprise (for example, AI that autocompletes reports, spreadsheet formulas and so on) ever be interoperable? Along with a coterie of organizations including Cloudera and Intel, the Linux Foundation — the nonprofit organization that supports and maintains a growing number of open source efforts — aims to find out.
The Linux Foundation on Tuesday announced the launch of the Open Platform for Enterprise AI (OPEA), a project to foster the development of open, multi-provider and composable (i.e. modular) generative AI systems. Under the purview of the Linux Foundation’s LF AI and Data org, which focuses on AI- and data-related platform initiatives, OPEA’s goal will be to pave the way for the release of “hardened,” “scalable” generative AI systems that “harness the best open source innovation from across the ecosystem,” LF AI and Data’s executive director, Ibrahim Haddad, said in a press release.
“OPEA will unlock new possibilities in AI by creating a detailed, composable framework that stands at the forefront of technology stacks,” Haddad said. “This initiative is a testament to our mission to drive open source innovation and collaboration within the AI and data communities under a neutral and open governance model.”
LF AI & Data Foundation Launches Open Platform for Enterprise AI (OPEA) for Groundbreaking Enterprise AI Collaboration
https://www.prnewswire.com/news-releases/lf-ai–data-foundation-launches-open-platform-for-enterprise-ai-opea-for-groundbreaking-enterprise-ai-collaboration-302117979.html
Open source solution empowers enterprises with open, multi-provider AI systems to drive unprecedented innovation and value for Generative AI
SEATTLE, April 16, 2024 /PRNewswire/ — Today, the LF AI & Data Foundation proudly announces the Open Platform for Enterprise AI (OPEA) as its latest Sandbox Project. This pioneering initiative unites industry leaders, including Anyscale, Cloudera, Datastax, Domino Data Lab, Hugging Face, Intel, KX, MariaDB Foundation, Minio, Qdrant, Red Hat, SAS, VMware (acquired by Broadcom), Yellowbrick Data, Zilliz and more to champion the development of open, multi-provider, robust, and composable GenAI systems.
The mission of LF AI & Data is to build and support an open artificial intelligence (AI) and data community and drive open source innovation in the AI and data domains by enabling collaboration and the creation of new opportunities for all members of the community. OPEA aims to facilitate and enable the development of flexible, scalable GenAI systems that harness the best open source innovation from across the ecosystem.
Tomi Engdahl says:
Copilot Enterprise now available
The world’s most widely adopted AI developer tool.
https://github.com/features/copilot
https://docs.github.com/en/copilot/quickstart
Is GitHub Copilot a free tool?
GitHub Copilot is free to use for verified students, teachers, and maintainers of popular open source projects. If you are not a student, teacher, or maintainer of a popular open source project, you can try GitHub Copilot for free with a one-time 30-day trial.
In order to get free GitHub Copilot, it’s only enough for you to be a student and sign up for GitHub Student Developer Pack. If you sign up for the GitHub Student Developer Pack, you can use so many other tools besides free GitHub Copilot.14 Nov 2023
Tomi Engdahl says:
https://www.zdnet.com/article/you-can-now-try-copilot-pro-for-free-and-heres-why-youll-want-to/
Can I get Copilot for free?
The free version of Microsoft Copilot already offers many features typically reserved for premium subscriptions, such as access to the internet and OpenAI’s most advanced large language model. If the free tier’s robustness has kept you from trying the premium plan, Copilot Pro, the following might change your mind.
Tomi Engdahl says:
Coding with an AI pair programmer: Getting started with GitHub Copilot
https://www.youtube.com/watch?v=dhfTaSGYQ4o
GitHub Copilot: the AI pair programmer for today and tomorrow
https://www.youtube.com/watch?v=AAT4zCfzsHI
Tomi Engdahl says:
https://www.neowin.net/news/microsoft-windows-copilot-app-we-quietly-installed-is-harmless-and-doesnt-steal-your-data/?fbclid=IwZXh0bgNhZW0CMTEAAR1rlxSFD3j7GeyK1EfoklKycB7QeMsw83UWd92XHd6lDi1UcrPifU2D5T8_aem_AaD_z3ZHay1Cs85gwH5xWSTXiFwJKiA1dNPrdyIt4QcEqYHKZTxCnLf01Ln6tE8_JdjdtMHVIDmepYqXbYLCL5EO
Tomi Engdahl says:
Molly White / Citation Needed:
An experiment with LLMs: the tools are helpful in mundane ways but you can’t build a hundred-billion-dollar industry around a technology that’s kind of useful
AI isn’t useless. But is it worth it?
https://www.citationneeded.news/ai-isnt-useless/
AI can be kind of useful, but I’m not sure that a “kind of useful” tool justifies the harm.
As someone known for my criticism of the previous deeply flawed technology to become the subject of the tech world’s overinflated aspirations, I have had people express surprise when I’ve remarked that generative artificial intelligence toolsa can be useful. In fact, I was a little surprised myself.
But there is a yawning gap between “AI tools can be handy for some things” and the kinds of stories AI companies are telling (and the media is uncritically reprinting). And when it comes to the massively harmful ways in which large language models (LLMs) are being developed and trained, the feeble argument that “well, they can sometimes be handy…” doesn’t offer much of a justification.
Some are surprised when they discover I don’t think blockchains are useless, either. Like so many technologies, blockchains are designed to prioritize a few specific characteristics (coordination among parties who don’t trust one another, censorship-resistance, etc.) at the expense of many others (speed, cost, etc.). And as they became trendy, people often used them for purposes where their characteristics weren’t necessary — or were sometimes even unwanted — and so they got all of the flaws with none of the benefits. The thing with blockchains is that the things they are suited for are not things I personally find to be terribly desirable, such as the massive casinos that have emerged around gambling on token prices, or financial transactions that cannot be reversed.
When I boil it down, I find my feelings about AI are actually pretty similar to my feelings about blockchains: they do a poor job of much of what people try to do with them, they can’t do the things their creators claim they one day might, and many of the things they are well suited to do may not be altogether that beneficial. And while I do think that AI tools are more broadly useful than blockchains, they also come with similarly monstrous costs.
Tomi Engdahl says:
Michael Peel / Financial Times:
Research suggests GPT-4 surpassed or achieved the same scores as all but top-scoring specialist medics in assessing ocular problems and suggesting treatments
OpenAI’s model all but matches doctors in assessing eye problems
Research harnesses power of generative artificial intelligence in latest sign of its potential use in medicine
https://www.ft.com/content/5b7a76be-467c-4074-8fd0-3e297bcd91d7
Tomi Engdahl says:
Bloomberg:
How GitHub Copilot became responsible for a significant percentage of coding, despite its limitations; Stack Overflow: 54.8% of developers used Copilot in 2023
Microsoft’s AI Copilot Is Starting to Automate the Coding Industry
https://www.bloomberg.com/news/articles/2024-04-17/microsoft-s-ai-copilot-is-starting-to-automate-the-coding-industry
The assistant is saving engineers hundreds of hours a month and is helping GitHub retain its edge over rivals including Amazon and Google.
When software developer Nikolai Avteniev got his hands on a preview version of Microsoft Corp.’s Copilot coding assistant in 2021, he quickly saw the potential.
Developed by Microsoft’s GitHub coding platform and based on a version of OpenAI’s generative artificial intelligence, the assistant wasn’t perfect and sometimes got things wrong. But Avteniev, who works for ticket seller StubHub, was surprised by how ably it finished lines of code with just a few prompts. All he had to do was
Tomi Engdahl says:
Samuel Stolton / Bloomberg:
Sources: the EU has decided that Microsoft’s OpenAI deal doesn’t merit a formal probe because it falls short of a takeover and Microsoft doesn’t control OpenAI — – Regulators don’t see evidence of Microsoft controlling OpenAI — Formal probe could have risked EU order to unwind the tie-up
https://www.bloomberg.com/news/articles/2024-04-17/microsoft-s-13-billion-openai-deal-to-avoid-formal-eu-probe
Tomi Engdahl says:
Ashley Belanger / Ars Technica:
The US NIST appoints as head of AI safety former OpenAI researcher Paul Christiano, who has been criticized for “AI doomer” views and ties to effective altruism — Former OpenAI researcher once predicted a 50 percent chance of AI killing all of us.
https://arstechnica.com/tech-policy/2024/04/feds-appoint-ai-doomer-to-run-us-ai-safety-institute/
Tomi Engdahl says:
Sheila Chiang / CNBC:
TSMC reports Q1 revenue up 16.5% YoY to NT$592.64B ($18.87B) vs. NT$582.94B est., profit up 8.9% YoY to NT$225.49B vs. NT$213.59B est., on strong AI chip demand
TSMC beats first-quarter revenue and profit expectations on strong AI chip demand
https://www.cnbc.com/2024/04/18/tsmc-q1-2024-earnings-on-strong-ai-chip-demand.html
Tomi Engdahl says:
Logitech toi ChatGPT:n hiireen ja näppäimistöön
https://etn.fi/index.php/13-news/16109-logitech-toi-chatgpt-n-hiireen-ja-naeppaeimistoeoen
PC-oheislaitteita valmistavalla Logitechilla joku keksi, että olisi hyvä idea, jos vaikkapa tekstiä käsitellessä saisi kutsuttua ChatGPT:n käyttöönsä yhdellä napin painalluksella. Siispä yhtiö kehitti Logi AI Prompt Builder -lisätoiminnon osaksi hiirien ja näppäimistöjen ohjausohjelmistoa.
Logi AI Prompt Builder on lisäikkuna, joka avautuu Logi Options+ -sovelluksen avulla yhden napin painalluksella. Se tuo tekoälyn käyttöön salamannopeasti. Käyttäjä voi esimerkiksi valita haluamansa tekstin ja antaa Logi AI Prompt Builderin muotoilla tekstin uudelleen, tehdä siitä yhteenveto tai kääntää toiselle kielelle.
Logi AI Prompt Builder on saatavana ilmaiseksi Windows- ja Mac-käyttäjille Logi Options+ -sovelluksen kautta. Sen käyttöliittymä toimii toistaiseksi vain englanniksi, mutta komentoja sille voi antaa myös suomeksi.
Tomi Engdahl says:
Ryan Morrison / Tom’s Guide:
Microsoft researchers introduce VASA-1, an AI model that creates a realistic talking face video from a portrait photo and an audio file, in research preview
Cool or creepy? Microsoft’s VASA-1 is a new AI model that turns photos into ‘talking faces’
https://www.tomsguide.com/ai/ai-image-video/microsoft-wants-your-photos-to-talk-vasa-1-is-a-new-ai-model-to-turn-images-into-talking-faces
Tomi Engdahl says:
Alex Heath / The Verge:
Meta details Llama 3: 8B- and 70B-parameter models, a focus on reducing false refusals, and an upcoming model trained on 15T+ tokens that has 400B+ parameters — Meta’s AI assistant is being put everywhere across Instagram, WhatsApp, and Facebook. Meanwhile, the company’s next major AI model, Llama 3, has arrived.
Meta’s battle with ChatGPT begins now
https://www.theverge.com/2024/4/18/24133808/meta-ai-assistant-llama-3-chatgpt-openai-rival
Meta’s AI assistant is being put everywhere across Instagram, WhatsApp, and Facebook. Meanwhile, the company’s next major AI model, Llama 3, has arrived.
New York Times:
Meta to add its Meta AI chatbot to Instagram, WhatsApp, Messenger, and Facebook in 12+ countries, expanding beyond the US, powered by its new Llama 3 models — Users of Instagram, Facebook, WhatsApp and Messenger will be able to turn to the new technology, powered by Meta’s latest artificial …
https://www.nytimes.com/2024/04/18/technology/meta-ai-assistant-push.html
Tomi Engdahl says:
Emma Roth / The Verge:
Meta rolls out real-time AI image generation, which changes the image as the user types a prompt, in beta on WhatsApp and Meta AI on the web in the US
https://www.theverge.com/2024/4/18/24134079/meta-real-time-ai-image-generation-whatsapp
Emilia David / The Verge:
Meta claims both Llama 3 models beat similarly sized models like Gemini, Mistral, and Claude 3 on certain benchmarks; humans marked Llama 3 higher than GPT-3.5
https://www.theverge.com/2024/4/18/24134103/llama-3-benchmark-testing-ai-gemma-gemini-mistral
Tomi Engdahl says:
David Pierce / The Verge:
Google announces a new Platforms and Devices team overseeing Pixel, Android, Chrome, Photos, and more, to integrate AI across its products, led by Rick Osterloh — AI is taking over at Google, and the company is changing in big ways to try to make it happen even faster.
https://www.theverge.com/2024/4/18/24133881/google-android-pixel-teams-reorg-rick-osterloh
Tomi Engdahl says:
https://www.uusiteknologia.fi/2024/04/19/tekoaly-auttaa-verkon-tekstinsyottoa-osaksi-kayttoliittymaa/
Tomi Engdahl says:
Five Eyes Agencies Release New AI Security Guidance
Five Eyes cybersecurity agencies have released joint guidance on securely deploying and operating AI systems.
https://www.securityweek.com/five-eyes-agencies-release-new-ai-security-guidance/
Tomi Engdahl says:
Emanuel Maiberg / 404 Media:
A look at Salad, which pays gamers in Fortnite skins and Roblox gift cards to rent their GPUs remotely to AI companies, including those making AI-generated porn
Gamers Are Renting Their Idle GPUs to Generate AI Porn
Emanuel Maiberg Emanuel Maiberg
·
Apr 18, 2024 at 9:15 AM
Salad, a company that pays gamers in Fortnite skins for their idle PCs, also generates AI porn.
https://www.404media.co/gamers-are-renting-their-idle-gpus-to-generate-ai-porn/