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,”
5,277 Comments
Tomi Engdahl says:
Deepa Seetharaman / Wall Street Journal:
OpenAI releases a tool to detect DALL-E 3-created images, claiming 98% accuracy for unaltered images, and joins Microsoft and Adobe’s content credentials group
OpenAI Says It Can Now Detect Images Spawned by Its Software—Most of the Time
Startup’s new tool detects 98% of pictures generated by its DALL-E 3 system, but success drops if the images are altered
https://www.wsj.com/tech/ai/openai-says-it-can-now-detect-images-spawned-by-its-softwaremost-of-the-time-83011149?st=hfix05fvfnpjfo1&reflink=desktopwebshare_permalink
AI is getting better at recognizing its own work.
OpenAI on Tuesday is launching a new tool that can detect whether an image was created using the company’s text-to-image generator, DALL-E 3. OpenAI officials said that the tool is highly accurate in detecting DALL-E 3 images, but that small changes to a picture can confuse it—reflecting how artificial-intelligence companies are playing catch up in the ability to track their own technology.
A surge of fake images and other media created using generative AI has created confusion about what is and isn’t real, and fueled discussion about the way images are affecting election campaigns in 2024.
Policymakers are concerned that voters are increasingly encountering AI-created images online and the wide availability of tools like DALL-E 3 make it possible to create such content even faster. Other AI startups and tech companies are also building tools to help.
“Election concern is absolutely driving a bunch of this work,” said David Robinson, who oversees policy planning for OpenAI. “It’s the number one context of concern that we hear about from policymakers.”
Tomi Engdahl says:
Mike Murphy / IBM Research:
IBM open sources its Granite code models for code generative tasks, trained on 116 programming languages, with models ranging in size from 3B to 34B parameters — Release — IBM is releasing a family of Granite code models to the open-source community. The aim is to make coding as easy …
https://research.ibm.com/blog/granite-code-models-open-source
Tomi Engdahl says:
Kyle Wiggers / TechCrunch:
Amazon launches Bedrock Studio in public preview, a web tool to help orgs experiment with and collaborate on generative AI models and then build AI-powered apps
Bedrock Studio is Amazon’s attempt to simplify generative AI app development
https://techcrunch.com/2024/05/07/bedrock-studio-is-amazons-attempt-to-simplify-generative-ai-app-development/
Amazon is launching a new tool, Bedrock Studio, designed to let organizations experiment with generative AI models, collaborate on those models, and ultimately build generative AI-powered apps.
Available in public preview starting today, the web-based Bedrock Studio — a part of Bedrock, Amazon’s generative AI tooling and hosting platform — provides what Amazon describes in a blog post as a “rapid prototyping environment” for generative AI.
Bedrock Studio guides developers through the steps to evaluate, analyze, fine-tune and share generative AI models from Anthropic, Cohere, Mistral, Meta and other Bedrock partners, as well as test different model settings and guardrails and integrate outside data sources and APIs. Bedrock Studio also offers tools to support collaboration with team members to create and refine generative AI apps, including single sign-on credentials for enterprises using them.
Bedrock Studio automatically deploys the relevant Amazon Web Services (AWS) resources as developers request them, Amazon says, and — in the interest of security — apps and data never leave the signed-in AWS account.
“When you create applications in Amazon Bedrock Studio, the corresponding managed resources such as knowledge bases, agents and [more] are automatically deployed in your AWS account,” Amazon principal developer advocate Antje Barth explains in the blog post. “You can use the Amazon Bedrock API to access those resources in downstream applications.”
Less an attempt to reinvent the wheel than streamline existing products and services, Bedrock Studio appears to be a stringing-together of AWS tools that have been around for some time, topped with a sprinkle of corporate governance and compliance capabilities. One imagines it’s all in service of Amazon’s bid to make Bedrock the go-to platform for generative AI app development.
https://aws.amazon.com/bedrock/studio/
Tomi Engdahl says:
Bryson Masse / VentureBeat:
Red Hat announces RHEL AI, a platform for developing and running open source LLMs, and InstructLab, a community project to let experts enhance AI models
Red Hat unveils RHEL AI and InstructLab to democratize enterprise AI
https://venturebeat.com/ai/red-hat-unveils-rhel-ai-and-instructlab-to-democratize-enterprise-ai/
At Red Hat Summit 2024 in Denver, Colorado, the open source software leader announced major new initiatives to bring the power of generative AI to the enterprise.
The headliners are Red Hat Enterprise Linux AI (RHEL AI), a foundation model platform for developing and running open source language models, and InstructLab, a community project to enable domain experts to enhance AI models with their knowledge.
Tomi Engdahl says:
Mitre Wants The Feds To Play In Its Sandbox
https://hackaday.com/2024/05/07/mitre-wants-the-feds-to-play-in-its-sandbox/
If you haven’t worked with the US government, you might not know Mitre, a non-profit government research organization. Formed in 1958 by the U.S. Air Force as a company to guide the SAGE computer, they are often research experts who oversee government contracts or evaluate proposals. Now they are building a $20 millon “AI Sandbox” for the Federal government to build AI prototypes.
Partnered with NVidia, the sandbox will use an NVidia GDX SuperPOD system capable of an exaFLOP of 8-bit AI computation. Mitre reports this will increase their compute power for AI by two orders of magnitude.
Access to the sandbox will be through one of the six federally funded R&D centers that Mitre operates on behalf of the government. These include centers that support the FAA, the IRS, Homeland Security, Social Security, health services, and cybersecurity with NIST. Of course, the DoD is likely in that mix, too.
Federal AI Sandbox
https://www.mitre.org/news-insights/fact-sheet/federal-ai-sandbox
To realize the incredible potential of AI within the federal government, a secure sandbox environment with significant computational power is needed for prototyping, training, and testing complex AI models.
Tomi Engdahl says:
Simon Willison / Simon Willison’s Weblog:
OpenAI built the gpt2-chatbot, renamed to “im-also-a-good-gpt-chatbot”, per the gpt2-chatbot’s 429 rate limit error message, which appeared in the LMSYS arena
https://simonwillison.net/2024/May/8/gpt2-chatbot-confirmed-as-openai/
Tomi Engdahl says:
Criminal Use of AI Growing, But Lags Behind Defenders
https://www.securityweek.com/criminal-use-of-ai-growing-but-lags-behind-defenders/
When not scamming other criminals, criminals are concentrating on the use of mainstream AI products rather than developing their own AI systems.
At the 2024 RSA Conference, taking place this week in San Francisco, Trend Micro on Wednesday delivered an update on its 2023 investigation into the criminal use of gen-AI. “Spoiler: criminals are [still] lagging behind on AI adoption.”
In summary, Trend Micro has found only one criminal LLM: WormGPT. Instead, there is a growing incidence, and therefore potential use, of jailbreaking services: EscapeGPT, BlackHatGPT, and LoopGPT. (The RSA presentation is supported by a separate Trend Micro blog.)
There is also an increasing number of ‘services’ whose purpose is unclear. These provide no demo and only mention their supposed capabilities: high on claims but low on proof. FraudGPT is one example.
Trend is not sure about the relevance or value of these offerings, and places them in a separate category labeled potential ‘scams’. Other examples include XXX.GPT, WolfGPT, EvilGPT, DarkBARD, DarkBERT, and DarkGPT.
In short, when not scamming other criminals, criminals are concentrating on the use of mainstream AI products rather than developing their own AI systems. This is also seen in the use of AI within other services. The Predator hacking tool includes a GPT feature using ChatGPT to assist scammers’ text creation abilities.
Tomi Engdahl says:
https://hackaday.com/2024/05/08/ai-helps-make-web-scraping-faster-and-easier/
Tomi Engdahl says:
Anna Tong / Reuters:
Sources: OpenAI plans to announce its AI-powered search product on May 13 — OpenAI plans to announce its artificial intelligence-powered search product on Monday, according to two sources familiar with the matter, raising the stakes in its competition with search king Google.
OpenAI plans to announce Google search competitor on Monday, sources say
https://www.reuters.com/technology/openai-plans-announce-google-search-competitor-monday-sources-say-2024-05-09/
May 9 (Reuters) – OpenAI plans to announce its artificial intelligence-powered search product on Monday, according to two sources familiar with the matter, raising the stakes in its competition with search king Google.
The announcement date, though subject to change, has not been previously reported. Bloomberg and the Information have reported that Microsoft (MSFT.O)
, opens new tab-backed OpenAI is working on a search product to potentially compete with Alphabet’s (GOOGL.O), opens new tab Google and with Perplexity, a well-funded AI search startup.
Tomi Engdahl says:
Ken Yeung / VentureBeat:
ElevenLabs previews a music-generating AI model, showing samples of songs with lyrics generated from text prompts — Voice AI startup ElevenLabs is offering an early look at a new model that turns a prompt into song lyrics. To raise awareness, it’s following a similar playbook Sam Altman used …
ElevenLabs previews music-generating AI model
https://venturebeat.com/ai/elevenlabs-previews-music-generating-ai-model/
Tomi Engdahl says:
Dan Milmo / The Guardian:
As part of its Model Spec release, OpenAI says it is considering letting developers and users “responsibly” generate “NSFW content in age-appropriate contexts” — Critics say ChatGPT creator’s proposal to allow erotica, slurs and other adult content undermines its mission statement
https://www.theguardian.com/technology/article/2024/may/09/openai-considers-allowing-users-to-create-ai-generated-pornography
Tomi Engdahl says:
Carl Franzen / VentureBeat:
OpenAI publishes Model Spec, which specifies how its AI models should act, including objectives, rules, and default behaviors, and asks the public for feedback
OpenAI posts Model Spec revealing how it wants AI to behave
https://venturebeat.com/ai/openai-posts-model-spec-revealing-how-it-wants-ai-to-behave/
Tomi Engdahl says:
Mark Stenberg / Adweek:
OpenAI’s pitch deck to news publishers promises priority placement to partners in chat conversations and payment contingent on “display success” — OpenAI’s Preferred Publisher Program offers media companies licensing deals — The generative artificial intelligence firm OpenAI …
Leaked Deck Reveals How OpenAI Is Pitching Publisher Partnerships
OpenAI’s Preferred Publisher Program offers media companies licensing deals
https://www.adweek.com/media/openai-preferred-publisher-program-deck/
Tomi Engdahl says:
Financial Times:
TikTok plans to start automatically labeling some AI content, including from DALL-E and Adobe’s Firefly, and joins the Adobe-led content credentials coalition
TikTok to automatically label AI-generated user content in global first
Social media companies under pressure to identify deepfakes and prevent spread of misinformation
https://www.ft.com/content/658c91f5-9eb2-4863-be0b-ee16ad1bc96c
Tomi Engdahl says:
Aisha Counts / Bloomberg:
Reddit releases a new content policy, including a ban on AI data licensees from using deleted posts or comments; Reddit expects $60M+ in 2024 licensing revenue
Reddit Lays Out Content Policy While Seeking More Licensing Deals
https://www.bloomberg.com/news/articles/2024-05-09/reddit-lays-out-content-policy-while-seeking-more-licensing-deals
Social media company restricts data use by AI companies
Reddit expects to generate $60 million from licensing in 2024
Tomi Engdahl says:
Emily Chang / Bloomberg:
An interview with Sundar Pichai on playing the long game on AI, how Google got Gemini image generation wrong, the future of search, layoffs, and more
Google’s Sundar Pichai Lays Out His AI Roadmap
https://www.bloomberg.com/news/articles/2024-05-08/video-alphabet-ceo-sundar-pichai-lays-out-google-s-ai-roadmap
Tomi Engdahl says:
EA is going all in on generative AI, will train models on 40 years of EA game data
EA CEO Andrew Wilson says developers have a ‘real hunger’ for generative AI, so expect the next Battlefield to include some form of AI
EA is going all in on generative AI, will train models on 40 years of EA game data
EA CEO Andrew Wilson says developers have a ‘real hunger’ for generative AI, so expect the next Battlefield to include some form of AI.
Read more: https://www.tweaktown.com/news/98152/ea-is-going-all-in-on-generative-ai-will-train-models-40-years-of-game-data/index.html?utm_source=dlvr.it&utm_medium=facebook&fbclid=IwZXh0bgNhZW0CMTEAAR2UKgIeqCD0HCPJy4u7N_tBWU-42QDoV64OvgJie7TBKTeCvRlWVuUWro4_aem_Adzck4Q9TEqLo_XwNxevg_7DA0FM35ZMIJ4oLJ-_PLwAMZJ4oJs1yqnXBcwzLgoLMrUGkWCBoCoL9aaHu27MJSLE
Tomi Engdahl says:
Big brains divided over training AI with more AI: Is model collapse inevitable?
Gosh, here’s us thinking recursion was a solved problem
https://www.theregister.com/2024/05/09/ai_model_collapse/
Tomi Engdahl says:
Stack Overflow bans users en masse for rebelling against OpenAI partnership — users banned for deleting answers to prevent them being used to train ChatGPT
News
By Dallin Grimm published 2 days ago
Stack Overflow is overflowing with salt.
https://www.tomshardware.com/tech-industry/artificial-intelligence/stack-overflow-bans-users-en-masse-for-rebelling-against-openai-partnership-users-banned-for-deleting-answers-to-prevent-them-being-used-to-train-chatgpt
Tomi Engdahl says:
https://yrityksille.elisa.fi/ideat/keskihintaiset-5g-puhelimet-ovat-houkutteleva-jarkiostos/
Tomi Engdahl says:
OpenAI destroyed a trove of books used to train AI models. The employees who collected the data are gone.
https://www.businessinsider.com/openai-destroyed-ai-training-datasets-lawsuit-authors-books-copyright-2024-5
Tomi Engdahl says:
https://www.euronews.com/culture/2024/04/25/miss-ai-worlds-first-beauty-contest-with-computer-generated-women
Tomi Engdahl says:
AI-generated beer commercial contains joyful monstrosities, goes viral
30-second spot set to “All Star” may inspire awe—or nightmares.
https://arstechnica.com/information-technology/2023/05/ai-generated-beer-commercial-contains-joyful-monstrosities-goes-viral/
Tomi Engdahl says:
OpenAI CEO Sam Altman promises “with a high degree of scientific certainty” that GPT-5 will be smarter than the “mildly embarrassing at best” GPT-4
News
By Kevin Okemwa published May 3, 2024
OpenAI’s GPT-5 model promises better performance and accuracy when using ChatGPT.
https://www.windowscentral.com/software-apps/openai-ceo-sam-altman-promises-gpt-5-will-be-smarter-than-gpt-4
Tomi Engdahl says:
Random robots are more reliable: New AI algorithm for robots consistently outperforms state-of-the-art systems
https://techxplore.com/news/2024-05-random-robots-reliable-ai-algorithm.html#google_vignette
Tomi Engdahl says:
Tekoäly on sitä mitä teemme siitä – ja samaan aikaan sekä uhka että mahdollisuus
Tekoäly on tällä hetkellä kuumaakin kuumempi aihe, ja sitä hyödynnetään monilla eri toimialoilla sisällöntuotannosta teollisuuteen asti. Yksi sen eniten keskustelua herättävä sovellusmuoto on kyberturvallisuus, jonka saralla tekoälyä voidaan hyödyntää niin puolustautumisessa kuin hyökkäämisessäkin. Miten tekoäly siis oikeastaan toimii ja mitkä ovat tärkeimmät asiat, jotka siitä pitäisi tietää?
https://www.dna.fi/yrityksille/blogi/-/blogs/tekoaly-on-sita-mita-teemme-siita-ja-samaan-aikaan-seka-uhka-etta-mahdollisuus?utm_source=facebook&utm_medium=linkad&utm_content=JUJO-artikkeli-tekoaly-on-sita-mita-teemme-siita-ja-samaan-aikaan-seka-uhka-etta-mahdollisuus&utm_campaign=H_JUJO_LAS_24-18-22_artikkelikampanja&fbclid=IwAR3ANMm2hAQ5vjraG8vJun6Nlz1H8TLo77NfKQwzklKJQaOb1WuAEzCytI8_aem_AbWDnOFJmBK47jNEloVidLI40mSXNdbrCBxixAyKT_t93YLR0YXQh1vaHiUafFgIH5e7B4cGMCPhIB-2Xwlr4P06
Tomi Engdahl says:
https://www.dna.fi/yrityksille/blogi/-/blogs/peter-nyman-johdon-on-luotsattava-organisaatio-tekoalyn-kayttoon?utm_source=facebook&utm_medium=linkad&utm_content=ALL-artikkeli-peter-nyman-johdon-on-luotsattava-organisaatio-tekoalyn-kayttoon&utm_campaign=T_ALL_ALL_24-16-19_yy_kevat&fbclid=IwAR0USNpegyqyQJDWtemx35wYtNyJLF8PakU-YxbOKV3iYXO868dXbq6fcts_aem_ARsIR3n7cVh5yMF-MiCQjEsNsLQ0G5JkI6FaWmwlsKBYheNfBE557vIoXF6xrc1xm6a8pjwnCgqh2pvEEuh5_NTG