Democratic operative admits to commissioning fake Biden robocall that used AI – https://www.nbcnews.com/politics/2024-election/democratic-operative-admits-commissioning-fake-biden-robocall-used-ai-rcna140402

Paper showing that ASCII art can get around AI guardrails. Its the return of 1980s hackers. https://t.co/1KGozsE4eQ https://t.co/zsDNBXqAFr” / X – https://twitter.com/emollick/status/1763687813386547594

Elon Musk sues OpenAI over AI threat | Courthouse News Service – https://www.courthousenews.com/elon-musk-sues-openai-over-ai-threat/

Read the lawsuit filing from Elon v. OpenAI
musk-v-altman-openai-complaint-sf.pdf – https://www.courthousenews.com/wp-content/uploads/2024/02/musk-v-altman-openai-complaint-sf.pdf

OpenAI accuses NYT of hacking ChatGPT to set up copyright suit | Ars Technica – https://arstechnica.com/tech-policy/2024/02/openai-accuses-nyt-of-hacking-chatgpt-to-set-up-copyright-suit/

New legal AI venture promises to show how judges think | Reuters –

A new venture by a legal technology entrepreneur and a former Kirkland & Ellis partner says it can use artificial intelligence to help lawyers understand how individual judges think, allowing them to tailor their arguments and improve their courtroom results.

https://www.reuters.com/legal/transactional/new-legal-ai-venture-promises-show-how-judges-think-2024-02-29

Malicious AI models on Hugging Face backdoor users’ machines – https://www.bleepingcomputer.com/news/security/malicious-ai-models-on-hugging-face-backdoor-users-machines/

“A GPT-4 powered customer service bot “is doing the equivalent work of 700 full-time agents” after a month. This is a press release from the company deploying it, so take it with a grain of salt, but it does match what I am seeing elsewhere and it suggests big implications. https://t.co/bmfQl677IO” / X – https://twitter.com/emollick/status/1762885468528533573

“AI: Consequential technology Karma says it’s AI assistant is doing the equivalent work of 700 full-time humans. Lays off those 700 humans. https://t.co/vweoSElCpC” / X – https://twitter.com/jowyang/status/1763152714194915774

“If you read Twitter discussions on AI you might think there is a lot of pushback to AI content. I am not sure that people are that discerning This experiment finds that people view AI content as just as reliable as human content, or as Wikipedia, and also prefer it stylistically https://t.co/5iJjiawqTK” / X – https://twitter.com/emollick/status/1762116887121371241

“openai legal entity structure is nuts https://t.co/2L7hkpDnqe” / X – https://twitter.com/iroasmas/status/1763947341038469418?s=20

Elon Musk sues OpenAI and CEO Sam Altman over contract breach – https://www.cnbc.com/2024/03/01/elon-musk-sues-openai-and-ceo-sam-altman-over-contract-breach.html

Klarna says OpenAI’s AI agent does the work of 700 support agents – https://the-decoder.com/klarna-says-openais-ai-agent-does-the-work-of-700-support-agents/

Klarna says its OpenAI virtual assistant does the work of 700 humans – https://www.fastcompany.com/91039401/klarna-ai-virtual-assistant-does-the-work-of-700-humans-after-layoffs

“Elon just sued OpenAI for breach of contract and fiduciary duty! He is claiming that OpenAI violated terms by becoming a closed source for profit corporation and is demanding that they open source their AI Don’t know if it will stand up in a legal court but it will… https://t.co/ZHRFqusA6N” / X – https://twitter.com/bindureddy/status/1763487614022324641

Scale AI to set the Pentagon’s path for testing and evaluating large language models | DefenseScoop – https://defensescoop.com/2024/02/20/scale-ai-pentagon-testing-evaluating-large-language-models/

“A reminder that even the best experts on LLMs don’t really understand all their behaviors (in this case, why AIs often seem a bit preachy & condescending). Lots left to learn.” / X – https://twitter.com/emollick/status/1763328898862121254

“Asking AIs for their political opinions is a hot topic, but this paper shows it can be misleading. LLMs don’t have them: “We found that models will express diametrically opposing views depending on minimal changes in prompt phrasing or situative context” https://t.co/krloCbr8lx https://t.co/2LYbrBXA5R” / X – https://twitter.com/emollick/status/1763022301514080592

“It is fascinating that, over a year after ChatGPT released, we still don’t actually know when, if, and how well frontier LLMs “reason.” There have been some papers suggesting limits to reasoning, but this new work finds that LLMs, while not as good as humans, can do reasoning.” / X – https://twitter.com/emollick/status/1762935652369248512

“”Do You Trust ChatGPT? – Perceived Credibility of Human and AI-Generated Content” participants did not substantially differentiate between Wikipedia-generated and LLM-generated content in terms of competence and trustworthiness. (Huschens et al, 2024) https://t.co/TKfaWqj73S https://t.co/UY5jrq2gZz” / X – https://twitter.com/WikiResearch/status/1762114374611710337

“An initial answer to the question I get asked most about prompting: “Should you be polite to the AI?” This cross-language study shows big LLMs give the best responses when you are moderately polite, but not obsequious. (I wish they had done more validation of prompt variants) https://t.co/lKF2acLxc7” / X – https://twitter.com/emollick/status/1761786151214440501

“Ironically, given the week of Twitter uproar about Gemini 1.0, Gemini 1.5 has one of the better approaches to dealing with risky content of the big LLMs I could see a slider system like this appearing elsewhere. It would be similarly easy to have them for other hot button topics https://t.co/1EKBcTJdBy” / X – https://twitter.com/emollick/status/1761249444375249040

*privacy not included | Shop smart and safe | Mozilla Foundation – https://foundation.mozilla.org/en/privacynotincluded/articles/happy-valentines-day-romantic-ai-chatbots-dont-have-your-privacy-at-heart/

“Lawsuits over AI training data won’t halt AI development, or even slow it much. This is for many reasons: open source models already released, a lack of consistent laws (training data isn’t protected in Japan) and the fact that many companies (Adobe) are now using licensed data.” / X – https://twitter.com/emollick/status/1764034777856790767

Klarna says its OpenAI virtual assistant does the work of 700 humans – https://www.fastcompany.com/91039401/klarna-ai-virtual-assistant-does-the-work-of-700-humans-after-layoffs

Google DeepMind forms a new org focused on AI safety | TechCrunch – https://techcrunch.com/2024/02/21/google-deepmind-forms-a-new-org-focused-on-ai-safety/

Hackers Could Use ChatGPT to Target 2024 Elections | TIME – https://time.com/6717129/hackers-ai-2024-elections/

Tumblr and WordPress to Sell Users’ Data to Train AI Tools – https://www.404media.co/tumblr-and-wordpress-to-sell-users-data-to-train-ai-tools/

Leo—Brave’s privacy-preserving AI assistant—now available on Android | Brave – https://brave.com/blog/leo-android/

Gemini Debacle

Google CEO tells employees Gemini AI blunder ‘unacceptable’ – https://www.cnbc.com/2024/02/28/google-ceo-tells-employees-gemini-ai-blunder-unacceptable.html

“A turning point of LLM deployment was when Microsoft effectively “posted through it.” After intense criticism of Bing kicked off by the Times article on how it acted like a stalker, Microsoft pulled down the AI… for two days. Then relaunched. Wonder if Google pushes through.” / X – https://twitter.com/emollick/status/1761492518653317584

Google Mired in Controversy Over Gemini AI Chatbot Push – WSJ – https://www.wsj.com/tech/ai/google-mired-in-controversy-over-ai-chatbot-push-46023dd3

Google CEO: Gemini AI photo diversity scandal ‘offended our users’ – The Verge – https://www.theverge.com/2024/2/28/24085445/google-ceo-gemini-ai-diversity-scandal-employee-memo

‘Embarrassing and wrong’: Google admits it lost control of image-generating AI | TechCrunch – https://techcrunch.com/2024/02/23/embarrassing-and-wrong-google-admits-it-lost-control-of-image-generating-ai/

Google aims to relaunch Gemini AI image tool in a few weeks – https://finance.yahoo.com/news/google-hopes-relaunch-gemini-ai-150314497.html

Be Sure To Read “This Week In AI”

This week’s executive overview and top links are here:

AI News #22: Week Ending 03/01/2024 with Executive Summary and Top 25 Stories

The post you just read is an extension of my weekly newsletter, This Week In AI, an executive summary of the top things to know in AI. Each week, I create an accessible overview for laypeople to feel confident they are conversant with the week’s AI developments. I include a curated list of must-click links of the week, to offer everyone a hands-on opportunity to explore the most intriguing updates in artificial intelligence across various categories, including robotics, imagery, video, AR/VR, science, ethics, and more. Beyond the overview, I post these topic-based deeper dives (below). If you haven’t read this week’s overview, I recommend starting there.

Credits/Sources

Most of these weekly links come from just a few prolific oversharing sources. Please follow them, as they work hard to find the news each week and they make it a lot easier for me to compile.