Tech billionaires talk to US Senate about AI - H+ Weekly - Issue #432
This week - Google nears release of Gemini; Apple spends millions per day on AI; e-skins and e-tattoos; old people are hot now; ChatGPT's traffic declines; and more!
On Wednesday, September 13 2023, US Senate Majority Leader Chuck Schumer hosted the first AI Insight Forum in Washington, DC. The purpose of these meetings is to assemble leading experts in artificial intelligence from various sectors, including business, academia, the arts, civil rights, and defence, to advance discussions on how Congress can address the challenges posed by AI.
For the inaugural AI Insight Forum, the senators invited the following people:
Sam Altman, CEO of OpenAI
Rumman Chowdhury, CEO of Humane Intelligence
Jack Clark, Co-founder of Anthropic
Clément Delangue, CEO of Hugging Face
Eric Fanning, President & CEO of the Aerospace Industries Association
Bill Gates, former CEO of Microsoft
Tristan Harris, Co-founder & ED of Center for Humane Technology
Jensen Huang, CEO of NVIDIA
Alex Karp, Co-founder & CEO of Palantir
Arvind Krishna, CEO of IBM
Janet Murguía, President of Unidos US
Elon Musk, CEO of X/Tesla
Satya Nadella, CEO of Microsoft
Sundar Pichai, CEO of Google
Deborah Raji, Researcher at U.C. Berkley
Charles Rivkin, Chairman & CEO of the Motion Picture Association
Eric Schmidt, former CEO of Google
Elizabeth Shuler, President of AFL-CIO
Meredith Stiehm, President of the Writers Guild
Randi Weingarten, President of the American Federation of Teachers
Maya Wiley, President & CEO of Leadership Conference on Civil & Human Rights
Mark Zuckerberg, CEO of Meta
Behind closed doors, over 60 US senators listened to what the invited guests had to say about AI and how it should be regulated. Listened, because the senators were not allowed to speak or even raise their hands. Since the meeting was held in private, we don’t know exactly what has been said, and our knowledge relies on statements made by attendees to the press after the session.
What we do know is that all the experts invited are in favour of government involvement in regulating AI. When Chuck Schumer, the forum's organizer, posed the question, "Does the government need to play a role in regulating AI?" every attendee raised their hand.
However, beyond this consensus, the meeting did not result in any concrete outcomes. Invited experts shared their views on AI and how it can be used for both good and bad. Elon Musk highlighted the “civilizational risks” of AI while Bill Gates focused on feeding the globe’s hungry. Some participants envisioned the creation of a comprehensive AI agency, while others argued that existing entities, such as the National Institute of Standards and Technology, were better suited to oversee AI regulation. There were also calls for urgency, with some arguing that the technology is moving fast and there is little time to put in place rules ensuring AI is safe.
One of the most contentious issues in the AI regulation discussion is Section 230 of the 1996 Communications Decency Act or its equivalent for AI. Section 230 provides legal immunity to social media platforms for content generated by users. Reports indicate that one panellist advised against a Section 230-like provision for AI and instead advocated for holding both users and technology creators accountable.
The meeting gathered some criticism. Some criticised the closed nature of the discussion, arguing that this conversation is too important to have it done behind closed doors. Others criticised the selection of experts invited to the panel, pointing out the overwhelming number of tech CEOs and the lack of people who actively research AI safety and the impact of technology on people’s lives. Some people point out that the tech CEOs publically say they are in favour of regulations while in the background they lobby to lax the rules.
So far, the US does not have a concrete proposal for how to regulate AI. We only have some initial proposals, like Senator Schumer’s SAFE framework. While the US policymakers engage in discussions with tech CEOs, academics and activists, the EU is moving forward with the AI Act and the Cyberspace Administration of China has proposed a concrete set of rules regarding AI. The conversation about how to regulate the AI industry in the US moves slowly compared to what the EU and China are doing. Thus far, the only thing the meetings hosted by the US Congress and the White House achieved is the commitment from the top US AI companies to AI safety (which is not legally binding or enforceable).
Becoming a paid subscriber is the best way to support the newsletter.
If you enjoy and find value in my writing, please hit the like button and share your thoughts in the comments. Share the newsletter with someone who will enjoy it, too.
You can also buy me a coffee if you enjoy my work.
🦾 More than a human
Old People Are Hot Now
This article makes an interesting observation - older people today are looking and feeling better than similarly aged people did 30 years ago. It makes a point that all those health guidelines popularised since the 1980s - eat healthy foods, exercise, stop smoking, etc. - seem to be working. This might also be a glimpse of what is coming when proper rejuvenation therapies become widely available, making our 50s and 60s not the beginning of the end but the beginning of the best years of our lives.
His Entire Body Was Paralyzed. Then They Put AI-Powered Chips in His Brain
Thanks to double neural bypass therapy, a man who was paralysed from the neck down was able to move and feel his body for the first time since the accident in 2020. The therapy involved embedding brain-computer interface chips into the man’s brain, electrode-laden patches on his spine and forearms and sensors in his fingertips and palms to bypass damaged nerves and restore the communication between his brain and the rest of the body. Since the procedure, the patient has regained full strength in both arms, with a 110% recovery in his right arm. He has also begun experiencing a natural recovery in his forearm and wrist, suggesting the therapy may stimulate the nervous system's healing processes.
▶️ E-skins and e-tattoos: Your cybernetic future (17:56)
Dr. Nanshu Lu works to create a future in which all those bulky medical devices and smartwatches designed to collect data about our bodies can replaced by soft, flexible and lightweight e-skin or e-tattoo devices. Her work also goes beyond medical applications and explores the possibility of using these devices to control robots or to use e-skin to give robots a sense of touch.
🧠 Artificial Intelligence
Google nears release of Gemini
Google has given a small group of companies access to an early version of Gemini, their highly-anticipated language model. Google is currently giving developers access to a relatively large version of Gemini, but not the largest version it is developing which would be more on par with GPT-4.
ChatGPT is losing some of its hype, as traffic falls for the third month in a row
Similarweb reports lower traffic to the ChatGPT website for the third month in a row. In June and July, the total number of visits declined by about 10%. August has seen a 3% decline. However, the traffic to ChatGPT might start to bounce back in September as people are coming back to school.
Apple is reportedly spending ‘millions of dollars a day’ training AI
Apple has silently joined the generative AI race. According to the report from The Information, Apple is spending millions of dollars per day on various AI projects, ranging from AI language models to text-to-image generators to multimodal AI. The report also mentioned Apple’s most advanced LLM, named Ajax. Ajax is said to be a 200-billion parameters model that is more powerful than OpenAI’s GPT-3.5. It’s fair to assume the first Apple product to use LLM will be Siri (we had a glimpse of that during this week’s iPhone event). I would not be surprised if machine learning engineers at Apple take more time to make the updated Siri powered by LLM run locally on the device and use that as a way to differentiate themselves from the competition from Google, OpenAI or Microsoft.
Adobe starts paying bonuses to Stock contributors whose content is being used to train Firefly
After months of testing, Adobe is rolling out AI tools to all Creative Cloud users. Alongside that announcement, Adobe said it will be paying artists whose work was used in training their models. All eligible Adobe Stock contributors with photos, vectors or illustrations in the standard and premium Stock collections will receive a bonus, which it plans to pay out annually going forward.
🤖 Robotics
Asteroid Landings Call for Robots With a Soft Touch
Landing a robot on an asteroid millions of kilometres away from Earth is a challenging task. A good example of what can go wrong here is the Philae lander which bounced off the surface of a comet. To avoid a similar failure in the future, some creative engineers look at soft robotics for inspiration and propose a shape-shifting articulated spacecraft that looks like a floating lily for future missions.
Biosignals, Robotics, and Rehabilitation
As society gets older, we might need more intelligent approaches to address healthcare challenges. Emerging technologies such as telemedicine, medical robots, powered prosthetics, exoskeletons, and AI-powered wearables might play a significant role here. This article focuses on a lab based at New York University Tandon School of Engineering that is working on exploring how to apply these technologies in medicine and how they can help people recovering from strokes and Covid-19, better predict disease spread and how telesurgery and telerobotic rehabilitation can be used in hospitals.
Necrobotics wins Ig Noble Prize
If you haven’t heard about the Ig Noble Prize yet, it is an annual award given to research that first makes you laugh and then makes you think. This year’s Mechanical Engineering Prize went to a team that used dead spiders as robot grippers thus creating a new field of research they called “necrobotics”.
🧬 Biotechnology
This “living material” self-destructs and cleans up polluted water
Researchers have developed a “living material” that could be used to clean up polluted water. The material itself consists of a seaweed-based gel packed with cyanobacteria genetically engineered to continually produce laccase - a protein that reacts with certain common water pollutants in a way that degrades them or makes them less toxic. In a proof of concept experiment, researchers showed that the bacteria are able to clean up indigo carmine, a toxic dye regularly used by the textile industry. Additionally, the bacteria were also genetically engineered to self-destruct in the presence of theophylline, a molecule found in black tea.
DARPA Seeks to SHIELD Blood from Fungal and Bacterial Pathogens
SHIELD (short for Synthetic Hemo-technologIEs that Locate & Disinfect) is DARPA’s new program which aims to develop new treatments for fungal and bacterial infections that can be a result of a gunshot or blast wounds, or burns. “SHIELD is designed to develop innovative approaches to create safe and effective broad spectrum medical countermeasures that can defeat fungal and bacterial pathogens, thereby preventing serious disease and death,” said SHIELD program manager, Dr. Christopher Bettinger.
H+ Weekly sheds light on the bleeding edge of technology and how advancements in AI, robotics, and biotech can usher in abundance, expand humanity's horizons, and redefine what it means to be human.
A big thank you to my paid subscribers, to my Patrons: whmr, Florian, dux, Eric, Preppikoma and Andrew, and to everyone who supports my work on Ko-Fi. Thank you for the support!
You can follow H+ Weekly on Twitter and on LinkedIn.
Thank you for reading and see you next Friday!