Last Week in AI #220: Neeva shuts down AI search, EU's proposed laws would ban open source LLMs, generative AI and the writer's strike, and more!

May 22, 2023

Next Steps for Neeva

Neeva, a search engine startup focused on using the latest advances in language models to provide a private, ad-free search experience, has announced that it will be shutting down its consumer search product due to difficulties in acquiring users and creating a sustainable business. The company had built an independent search stack and integrated generative AI and LLMs to create an answer engine. However, convincing users to switch from their current search engines proved to be a major challenge. Neeva will now shift its focus to applying its search and LLM expertise in enterprise settings.

Opinion/Analysis: The shutdown of Neeva's consumer search product highlights the immense challenge of competing against established search engines like Google. Despite offering a unique value proposition, Neeva faced an uphill battle in convincing users to switch from their default search engines. This raises questions about the potential for new entrants in the search engine market, as it’s much easier for incumbents to integrate AI into existing products than for new companies to poach existing users.

EU AI Act To Target US Open Source Software

The EU's amended AI Act aims to ban American companies like OpenAI, Amazon, Google, and IBM from providing API access to generative AI models in Europe. The act would impose massive fines on companies and open-source developers if unlicensed generative models become available in the region. The legislation has a broad jurisdiction, requires registration and expensive risk testing for high-risk AI projects, and has vaguely defined risks. Open-source generative AI models are not exempt, and API access is essentially banned. The act also has extraterritorial jurisdiction, allowing European governments to seek conflict with American developers and businesses.

Opinion/Analysis: While the AI Act has not been passed yet (and the content of the legislation is likely to change), this current iteration appears to be a misguided attempt to regulate AI technology, with potentially negative consequences for innovation and the tech ecosystem. By targeting open-source developers and imposing strict regulations on API access, the act could stifle creativity and hinder the growth of small businesses. This may signal to AI developers that the EU is not a friendly market for AI applications, which will certainly backfire on AI development and adoption in the EU.

Why use of AI is a major sticking point in the ongoing writers' strike

People picket outside of FOX Studios on the first day of the Hollywood writers strike on May 2, 2023 in Los Angeles

The Writers Guild of America (WGA) strike has brought attention to the use of AI in the film and television industry, with the union proposing a ban on AI-generated story ideas and scripts. The WGA also wants to prevent scripts covered by their collective bargaining agreement from being used to train AI models. Hollywood studios and streaming services have rejected the proposal, offering annual meetings to discuss technological changes instead. The strike highlights concerns about AI's potential impact on writers' compensation, writing credits, and the devaluation of their craft.

Opinion/Analysis: The WGA strike raises important questions about the role of AI in creative industries and the potential consequences for human workers. The WGA's demands signal a growing awareness among creative professionals that AI could threaten their livelihoods and the value of their work. The outcome of this strike could set a precedent for how other creative industries approach the integration of AI and the protection of human workers. It's crucial for both sides to find a balance between embracing technological advancements and preserving the unique contributions of human creativity.

This Issue is Sponsored by: The No Priors Podcast

The No Priors podcast is your guide to the AI revolution. At this moment of inflection in technology, co-hosts Elad Gil and Sarah Guo ask the world's leading AI engineers, researchers, and founders the biggest questions: How far away is AGI? What markets are at risk for disruption? How will commerce, culture, and society change? What’s happening in state-of-the-art research? Listen here

Other News

Research

New AI research lets you click and drag images to manipulate them in seconds - "No, it’s not over yet: the ability of AI tools to manipulate images continues to grow. The latest example is only a research paper for now, but a very impressive one, letting users simply drag elements of a picture to change their appearance."

Microsoft AI Research Introduces Automatic Prompt Optimization (APO): A Simple and General-Purpose... - "APO is a general and nonparametric prompt optimization algorithm inspired by numerical gradient descent."

AI Scientists: Safe and Useful AI? - "they are modeled after ideal scientists and do not act autonomously in the real world, only focusing on theory building and question answering."

Can't find your phone? There's a robot for that - "Robots can help find objects you've lost, thanks to new 'artificial memory'"

Better than humans: Artificial intelligence in intensive care units - "An artificial intelligence developed at TU Wien (Vienna) can suggest appropriate treatment steps in cases of blood poisoning. The computer has already surpassed humans in this respect."

Machine learning model able to detect signs of Alzheimer's across languages - "The model was able to distinguish Alzheimer's patients from healthy controls with 70 to 75 percent accuracy, a promising figure for the more than 747,000 Canadians who have Alzheimer's or another form of dementia."

Machine learning can more accurately predict ocean currents - "To investigate ocean currents, scientists usually release GPS-tagged buoys and record their velocities to reconstruct the currents that carry them."

Beyond the Spectrum: Machine Learning Unlocks Predictive Power in Organic Chemistry Research - "This breakthrough, trained on a database of over 22,000 molecules, could accelerate the design of functional molecules like pharmaceuticals."

Meet VideoChat: An End-to-End Chat-Centric Video Understanding System Developed by Merging Language and Visual Models - "Real-world applications like autonomous driving and human-robot interaction rely heavily on intelligent visual understanding."

Machine learning reveals sex-specific Alzheimer's risk genes - "A new study in the journal Nature Communications discusses the possibility of detecting genetic variants that could predict the risk of Alzheimer’s disease (AD) in males and females using a machine learning (ML) approach."

Applications

Apple’s new ‘Personal Voice’ feature can create a voice that sounds like you or a loved one in just 15 minutes - "Personal Voice, on the other hand, is a way for people who are at risk of losing their ability to speak to create and save a voice that sounds like them. "

Deal Dive: AI relationship coach Amorai offers more questions than answers - "The startup has built an AI relationship coach to help people grow and foster real-life connections by offering advice and answers to relationship questions. "

Hybe uses AI voice technology for Midnatt's multilingual single - "K-pop powerhouse Hybe on Monday unveiled its first artificial intelligence (AI) project Midnatt, which sings a song in multiple languages and switches between female and male voices using generative voice technology."

Machine learning algorithm a fast, accurate way of diagnosing heart attack - "The machine learning algorithm was "taught" using the patient’s sex, age, ECG findings and medical history, in addition to troponin levels, to identify the probability that a heart attack had occurred."

AI-powered coding, free of charge with Colab - "Google Colab will soon introduce AI coding features using Google’s most advanced family of code models, Codey."

llm, ttok and strip-tags—CLI tools for working with ChatGPT and other LLMs - "I’ve been building out a small suite of command-line tools for working with ChatGPT, GPT-4 and potentially other language models in the future. "

Audio journalism app Curio can now create personalized episodes using AI - "Curio, a startup building a platform that turns expert journalism into professionally narrated content, is embracing AI technology to create customized audio episodes, based on your prompts. "

The Daring Robot Surgery That Saved a Man’s Life - "Two doctors, separated by thousands of miles, carried out a lifesaving operation using a robot. It’s the start of a major change in how surgery is performed."

Business

Phoenix is the world’s first humanoid general-purpose robot powered by Carbon.

Meet Phoenix, a general-purpose humanoid robot designed for work - "Canada’s Sanctuary AI, which aims to create the world’s-first human-like intelligence in general-purpose robots, has now announced a major step forward with the unveiling of its sixth-generation general-purpose robot named Phoenix."

VCs love to talk about AI, but they aren’t writing as many checks as you might think - "CB Insights data shows pronounced slowdown in AI investment"

OpenAI readies new open-source AI model - The Information - "OpenAI is preparing to release a new open-source language model to the public, The Information reported on Monday, citing a person with knowledge of the plan."

Together raises $20M to build open source generative AI models - "“Together is spearheading AI’s ‘Linux moment’ by providing an open ecosystem across compute and best in class foundation models,”"

Hippocratic is building a large language model for healthcare - "AI, specifically generative AI, has the potential to transform healthcare. At least, that sales pitch from Hippocratic Health, which emerged from stealth today with a whopping $50 million in seed financing behind it and a valuation in the “triple digit millions."

Amazon Plans to Add ChatGPT-Style Search to Its Online Store - "Company’s AI ambitions appear in job listings for engineers"

Introducing the ChatGPT app for iOS - "The ChatGPT app syncs your conversations, supports voice input, and brings our latest model improvements to your fingertips."

Would You Take Financial Advice From A.I.? - "The financial services industry is plotting how to incorporate tools like ChatGPT into its products. But humans will still be necessary to provide personal advice."

Cambridge-based TidalSense raises €8.6 million to spread its AI-driven tech for respiratory conditions - "The Cambridge-based AI-driven company has created a handheld medical device (N-Tidal) that detects changes in lung function sensitively and enables quicker, more accurate and automated diagnosis of chronic obstructive pulmonary disease (COPD), with asthma to follow. "

BT Chief Says AI Linked to 10,000 of Total Planned Job Cuts - "CEO Jansen said AI will help BT in customer service, networks"

Nvidia Chip King's Fortune Doubles This Year to $27.3 Billion on AI Boom - "Jensen Huang's wealth has grown at a faster rate in 2023 than any other US billionaire."

Meta pulls the curtain back on its A.I. chips for the first time - "Meta has built custom computer chips to help with its artificial intelligence and video-processing tasks and is talking about them in public for the first time."

Latest moves show Stability AI is fully committed to open source — well, mostly - "Today, Stability AI released StableStudio, an open-source version of DreamStudio, its commercial interface for generating Stable Diffusion images."

Meta AI's contribution to the 'Speed of A.I.' is Considerable - "Open-source, Chips, A.I. datacenters collide at Meta."

Study: 47% of Americans have used ChatGPT for stock recommendations - "ChatGPT has shown promise as a tool for investors, but it’s not ready to replace your financial advisor."

Google plans to use new A.I. models for ads and to help YouTube creators, sources say - "Google's effort to rapidly add new artificial intelligence technology into its core products is making its way into the advertising world, CNBC has learned."

StableStudio is Stability AI’s latest commitment to open-source AI - "Stability AI is releasing an open-source version of DreamStudio, a commercial interface for the company’s AI image generator model, Stable Diffusion."

Another Side of the A.I. Boom: Detecting What A.I. Makes - "More than a dozen companies have popped up to offer services aimed at identifying whether photos, text and videos are made by humans or machines."

Concerns

The Alan Turing Institute has failed to develop modern AI in the UK - "The UK’s flagship institute for artificial intelligence, the Alan Turing Institute, has been at best irrelevant to the development of modern AI in the UK."

Tesla’s “Full Self-Driving” sees pedestrian, chooses not to slow down - "Fan is gleeful about video of beta Full Self-Driving mode breaking traffic laws."

Google tool to spot fake AI images after puffer jacket Pope went viral - "About This Image search tool will help people tell hyper realistic pictures from real ones"

Not even NYT bestsellers are safe from AI cover art - "A prominent fantasy novel features a cover apparently generated with artificial intelligence, sparking complaints from artists and book enthusiasts. Earlier this month, readers noted that the back of the UK edition of Sarah J."

Poll: 61% of Americans say AI threatens humanity’s future - "New Reuters/Ipsos poll reveals religious and political divides over AI as well."

Texas A&M Prof Flunks All His Students After ChatGPT Falsely Claims It Wrote Their Papers - "Texas A&M University–Commerce seniors who have already graduated were denied their diplomas because of an instructor who incorrectly used AI software to detect cheating"

Elon Musk’s letter to Microsoft CEO: You misused Twitter data - "Twitter’s API-driven apps like Hootsuite, Spriklr and Sprout Social were to free to use for some researchers, partners and developers who agreed to their terms, reported CNBC."

Don’t Go Down That AI Longtermism Rabbit Hole - "Silicon Valley is fixated on longtermism — prioritizing future lives over those living today. This is very tricky moral math."

CNET Journalists Seek to Unionize, Saying AI ‘Threatens Our Jobs and Reputations’ - "Bargaining unit aims to represent a staff of about 100"

Baidu boss charged with stealing Apple’s self-driving car tech - "A former Apple engineer and the executive at an electric vehicle startup owned by the Chinese tech giant Baidu has been charged with stealing Apple’s self-driving technology, as reported earlier by CNBC."

Why AI’s diversity crisis matters, and how to tackle it - "Machine-learning researchers from under-represented groups say the field desperately needs more people like them to ensure the technologies deliver for all."

6 major risks of using ChatGPT, according to a new study - "For all of the good things you can do with OpenAI's new chatbot, you also need to be aware of the ways it could be used by people with malicious intent."

Age of Danger' explores potential risks because AI doesn't understand rules of war - "NPR's Steve Inskeep speaks with Thom Shanker, co-author of the book Age of Danger, about the threats artificial intelligence poses to national security."

Analysis

That wasn’t Google I/O — it was Google AI - "If you thought generative AI was a big deal last year, wait until you see what it looks like in products already used by billions."

The apocalypse isn’t coming. We must resist cynicism and fear about AI - "Remember when WeWork would kill commercial real estate? Crypto would abolish banks? The metaverse would end meeting people in real life?"

Why No One Wants AI Ethics Anymore - "And what they should do to recover the credibility they've lost"

Just Calm Down About GPT-4 Already - "And stop confusing performance with competence, says Rodney Brooks"

Unfortunately, OpenAI and Google have moats - "The companies that have users interacting with their models consistently have moats through data and habits."

Diagram of an AI Image System - "With the understanding that there is no such thing as an "AI image""

The Optimist’s Guide to Artificial Intelligence and Work - "The focus of much discussion is on how it will replace jobs, but nothing is inevitable."

LLMemento - "A striking thing about large language models and diffusion models — what we talk about nowadays, when we say “AI” — is that they have no memory. Zero. None. They are exactly the same system, doing exactly the same thing, every single time."

Policy

The open-source AI boom is built on Big Tech’s handouts. How long will it last? - "Greater access to the code behind generative models is fueling innovation. But if top companies get spooked, they could close up shop."

OpenAI Tells Congress the U.S. Should Create AI 'Licenses' to Release New Models - "Many AI researchers see this as an anti-competitive move, as requiring licensing will be much more beneficial for larger companies than smaller ones."

Washington is struggling to catch up on artificial intelligence - "As industry experts and academics sound alarm bells, Washington has quickly turned its attention to artificial intelligence. But there’s hardly a coordinated approach when it comes to regulating the fast-moving technology."

Labour should pledge £11bn to build ‘BritGPT’ AI, thinktank says - "Keir Starmer should pledge £11bn towards building “BritGPT” and a national artificial intelligence (AI) cloud in the next Labour manifesto or risk the UK falling ever further into dependence on American tech companies, an affiliated thinktank has said."

US Senator Uses ChatGPT for Opening Remarks at a Hearing on AI - "US Senator Richard Blumenthal opened a hearing on AI with a recording of his voice describing the risks of this emerging technology."

US Senator Richard Blumenthal opened a hearing on AI with a recording of his voice describing the risks of this emerging technology. - "Eric Schmidt to testify Wednesday at House hearing on issue"

NYC Public Schools Drop Ban on AI Tool ChatGPT - "New York City public schools lifted a ban on artificial intelligence tool ChatGPT after initially banning the technology."

Apple restricts use of OpenAI's ChatGPT for employees -WSJ - "Apple Inc (AAPL.O) has restricted the use of ChatGPT and other external artificial intelligence tools for its employees as Apple develops similar technology, the Wall Street Journal reported on Thursday, citing a document and sources."

Bipartisan Senate group looks to tackle AI legislation - "A bipartisan group of senators is working on regulation for artificial intelligence as lawmakers consider the best avenue for passing legislation."

In Battle Over A.I., Meta Decides to Give Away Its Crown Jewels - "The tech giant has publicly released its latest A.I. technology so people can build their own chatbots. Rivals like Google say that approach can be dangerous."

Italy earmarks $33M to protect workers from AI replacement threat - "Futurist and WAYE founder Sinead Bovell joined ‘Mornings with Maria’ to discuss a few of the dangerous aspects of advancing artificial intelligence technology. Per the Fondo per La Repubblica Digitale (FRD), the funding for digital training will be allocated in two ways."

Last Week in AI