A Race to Extinction: How Great Power Competition Is Making Artificial Intelligence Existentially Dangerous
...
Beyond the world of text, generative applications Midjourney, DALL-E, and Stable Diffusion produceunprecedentedly realistic images and videos. These models have burst into the public consciousness rapidly. Most people have begun to understand that generative AI is an unparalleled innovation, a type of machine that possesses capacities — natural language generation and artistic production — long thought to be sacrosanct domains of human ability.
But generative AI is only the beginning. A team of Microsoft AI scientists recently released a paperarguing that GPT-4 — arguably the most sophisticated LLM yet — is showing the “sparks” of artificial general intelligence (AGI), an AI that is as smart — or smarter — than humans in every area of intelligence, rather than simply in one task. They argue that, “[b]eyond its mastery of language, GPT-4 can solve novel and difficult tasks that span mathematics, coding, vision, medicine, law, psychology and more, without needing any special prompting." In these multiple areas of intelligence, GPT-4 is “strikingly close to human-level performance.” In short, GPT-4 appears to presage a program that can think and reason like a human. Half of surveyed AI experts expect an AGI in the next 40 years. ...
The pressure to outpace adversaries by rapidly pushing the frontiers of a technology that we still do not fully understand or fully control — without commensurate efforts to make AI safe for humans — may well present an existential risk to humanity’s continued existence. ...
...just the perception of falling behind an adversary contributed to a destabilizing buildup of nuclear and ballistic missile capabilities, with all its associated dangers of accidents, miscalculations, and escalation. ...
The Alignment Problem
Despite dramatic successes in AI, humans still cannot reliably predict or control its outputs and actions. While research focused on AI capabilities has produced stunning advancements, the same cannot be said for research in the field of AI alignment, which aims to ensure AI systems can be controlled by their designers and made to act in a way that is compatible with humanity’s interests. ...
Arms Racing or Alignment Governance? A Risky Tradeoff
How does international competition come into play when discussing the technical issue of alignment? Put simply, the faster AI advances, the less time we will have to learn how to align it. ...
Likewise, the perception of an arms race may preclude the development of a global governance framework on AI. A vicious cycle may emerge where an arms race prevents international agreements, which increases paranoia and accelerates that same arms race. ...
However, the outlook is not all rosy: as the political salience of AI continues to increase, the questions of speed, regulation, and cooperation may become politicized into the larger American partisan debate over China. Regulation may be harder to push when “China hawks” begin to associate slowing AI with losing an arms race to China. Recent rhetoric in Congress has emphasized the AI arms race and downplayed the necessity of regulation. ...
See the full story here: https://hir.harvard.edu/a-race-to-extinction-how-great-power-competition-is-making-artificial-intelligence-existentially-dangerous/
TIME Reveals Inaugural TIME100 AI List of the World’s Most Influential People in Artificial Intelligence
See the list here: https://www.prnewswire.com/news-releases/time-reveals-inaugural-time100-ai-list-of-the-worlds-most-influential-people-in-artificial-intelligence-301920536.html
[McKinsey] What’s the future of generative AI? An early view in 15 charts
See the full story with 15 charts here: https://www.mckinsey.com/featured-insights/mckinsey-explainers/whats-the-future-of-generative-ai-an-early-view-in-15-charts?cid=other-eml-shl-mip-mck&hlkid=4a4d7c3a72aa49968c53834a9cccf217&hctky=2717270&hdpid=4a4b61b4-1426-42b1-aa20-8e6948de518d
The Role of Reputation Systems Within Decentralized Autonomous Organizations
Decentralized autonomous organizations (DAOs) have become an increasingly popular way to organize groups and projects in a decentralized manner through blockchain technology. DAOs allow people to collaborate and make decisions collectively without centralized control. However, ensuring accountability within decentralized structures presents unique challenges. This is where reputation systems come in.
Reputation systems help build trust and alignment among members of a decentralized community. Within a DAO, reputation systems allow members to rate each other based on contributions, character, and reliability. High reputation scores act as a signal of status and trustworthiness. In a space where anonymity and pseudonymity are common, reputation systems create transparency around identity and track record. As DAOs continue to gain adoption, well-designed reputation systems will be key to their success. ...
Members are incentivized to contribute value and act honestly to garner high reputation scores. Concurrently, high reputation members signal that they are trustworthy allies. ...
Mentorship programs matching high reputation members with newcomers also foster inclusive onboarding. Temporary reputation boosts for being vouched by existing members is another onboarding design. The goal is healthy socialization and information osmosis. ...
See the full story here: https://btcpeers.com/the-role-of-reputation-systems-within-decentralized-autonomous-organizations/
Google to require disclosure for AI in election ads
...
Why it matters: The explosion of generative AI has raised questions over its use to create deceptive content in an online political environment already rife with misinformation.
- Online election ads remains a fairly unregulated space, though the FEC is mulling expanded new rules, as Axios previously reported. ...
Details: Language on the disclosures may say things like "this audio was computer generated" or "this video content was synthetically generated," per the announcement. ...
- The policy will be enforced using a combination of human review and tech tools to examine ads, and ads will be disapproved or removed accordingly. Advertiser accounts that engage in pervasive or egregious violations may be suspended, Aciman said.
- Advertisers will be notified if an ad is rejected and will have the ability to appeal and re-upload with the proper disclosure.
See the full story here: https://www.axios.com/2023/09/06/google-ai-election-ads-disclosure
You Are Here – by Evan Shapiro – Media War & Peace
... The last Media era ended in 2019 when Disney+ launched. That coincided with Covid. So, the actual onset of the new, current era was obscured by Media’s explosive and irrational growth during lockdown. At the end of the last era, the moguls all went all-in on business models they knew nothing about. As Hollywood distractedly chased Netflix and Spotify down their rabbit holes, big tech captured even more share of the world’s hearts, minds and pockets. Now, last era’s moguls are trying to cut their way to success, like its 2008. ...
Media platforms designed for this new User-Centric era already merge content and technology into hyper-personalized lifestyle bundles. Media businesses made for this moment, already operate on Multi-faceted Models, across the hierarchy of consumer feeds. ...
See the full story here: https://eshap.substack.com/p/you-are-here?r=5xhla&utm_source=substack&utm_medium=email
Exclusive survey: Experts favor new U.S. agency to govern AI
See the complete story here: https://www.axios.com/2023/09/05/ai-regulations-expert-survey
The Fate of Apple’s Vision Pro: Part 1 and 2
PhilNote: This is a very good combination of observation and speculation about the Apple Vision Pro and Spatial Computing.
Part 1
https://www.mediumenergy.io/p/why-apples-vision-pro-matters-part
Part 2
... Unlike other players in this space, most people will give Apple the benefit of the doubt.
More so than perhaps any other company, Apple knows how to make things desirable. Which is a key pillar of their strategy: social engineering. They’re going to make this thing cool and they have a plan to do so. One example is hyper personalization.
Apple is going to ensure your AVP fits like a glove, while also offering the opportunity for self-expression and style via custom aesthetics....
But if 'isolation' is your concern, know that collaboration will be the killer feature of spatial computing’s killer apps. ...
Across all of these examples, people are fundamentally trying to transfer knowledge by conveying an 'experience' in a woefully non-experiential way. This ‘knowledge transfer’ problem becomes increasingly acute in the face of an aging workforce, worker displacement (as AI eats more jobs), and within an era of customization/personalization in product design & sales.
This is why I prefer to call spatial computing, 'experiential computing'....
Apple buys Swedish classical record label
- BIS Records founder Robert von Bahr said Tuesday that Apple has bought the classical music-focused record company and will fold it into its Apple Music Classical and Platoon music services.
- The acquisition is another example of Apple attempting to distinguish itself from streaming music rivals like Spotify by focusing on classical music.
- The tech giant debuted the Apple Music Classical streaming app this past spring, pitching it as a way for subscribers to access over 5 million classical tracks that can be searched via data like composer, conductor or catalog number.
See the full story here: https://www.cnbc.com/2023/09/05/apple-buys-swedish-classical-record-label.html
Ban or Embrace? Colleges Wrestle With A.I.-Generated Admissions Essays.
Rick Clark, the executive director of undergraduate admission at the Georgia Institute of Technology, and his staff spent weeks this summer pretending to be high school students using A.I. chatbots to fill out college applications....
“Students on some level are going to have access to and use A.I.,” Mr. Clark said. “The big question is: How do we want to direct them, knowing that it’s out there and available to them?” ...
“Part of the process of the college essay is finding your writing voice through all of that drafting and revising,” said Susan Barber, an Advanced Placement English literature teacher at Midtown High School, a public school in Atlanta. “And I think that’s something that ChatGPT would be robbing them of.”
In August, Ms. Barber assigned her 12th-grade students to write college essays. This week, she held class discussions about ChatGPT, cautioning students that using A.I. chatbots to generate ideas or writing could make their college essays sound too generic. She advised them to focus more on their personal views and voices.
Other educators said they hoped the A.I. tools might have a democratizing effect. Wealthier high school students, these experts noted, often have access to resources — alumni parents, family friends, paid writing coaches — to help them brainstorm, draft and edit their college admissions essays. ChatGPT could play a similar role for students who lack such resources, they said, especially those at large high schools where overworked college counselors have little time for individualized essay coaching. ...
Several high school seniors said in interviews that they had chosen not to use A.I. tools to help draft their essays — partly because they wanted to tell their own personal stories themselves, and partly because many universities have not taken clear stances on applicants’ use of the chatbots. ...
Ritika Vakharia, a senior at the Gwinnett School of Mathematics, Science and Technology, said she had tried asking ChatGPT to produce ideas for college admissions essays. But she found the responses too broad and impersonal, even after she gave it details about her extracurricular activities like teaching dance classes to younger students. ...
See the full story here: https://www.nytimes.com/2023/09/01/business/college-admissions-essay-ai-chatbots.html
Pages
- About Philip Lelyveld
- Mark and Addie Lelyveld Biographies
- Presentations and articles
- Trustworthy AI – A Market-Driven approach
- Tufts Alumni Bio