... “It really shows that these models are not going to be reliable in the long run if they learn from post-LLM age data—without being able to tell what data has been machine-generated, the quality of the outputs will continue to decline,” says Catherine Flick, a professor of ethics and games technology at Staffordshire University.
The reason for that decline is the recursive nature of the LLM loop—and exactly what could have caused the snafu with Grok. ...
Winterbourne’s issues with Grok are just the tip of the iceberg. A visual representation of the damage that model collapse can have has been demonstrated by researchers at Stanford University and the University of California, Berkeley, who fed generative AI image creators with AI-generated output. The distortions and warping that happened turned perfectly normal human faces into grotesque caricatures, as the model begins to break. The fun “make it more” meme that has circulated on social media, where users ask AI image generators to make their output more extreme, also highlights what can happen when AI begins to train itself on AI-generated output. ...
“Each generation of a particular model will be that much less reliable as a source of true facts about the world because each will be trained with an ever less reliable data set,” says Mike Katell, ethics fellow at the Alan Turing Institute. “ ...
... Altman admits that there are challenges that demand close attention. One particular concern to be wary of, with 2024 elections on the horizon, is how AI stands to influence democracies. Whereas election interference circulating on social media might look straightforward today—“troll farms…make one great meme, and that spreads out”—Altman says that AI-fueled disinformation stands to become far more personalized and persuasive: “A thing that I’m more concerned about is what happens if an AI reads everything you’ve ever written online … and then right at the exact moment, sends you one message customized for you that really changes the way you think about the world.” ...
Bose is discontinuing its Frames line of audio smart glasses, which will no longer be available by the end of this year. ...
The first iteration of Bose Frames was pitched as a pair of smart glasses with audio-based augmented reality — a showcase for Bose AR. You could receive GPS audio directions, for example, and there was a smattering of audio-based story games. ...
It’s even more disappointing because this past year has seen a mini-resurgence in smart glasses with the second-gen Ray-Ban Meta smart glasses and the third-gen Amazon Echo Frames. However, both those glasses do more than the Bose Frames. ...
The Bose Frames may have had good audio, but that alone simply wasn’t enough.
... Apple rolled out iOS 17.2 today, giving iPhone 15 Pro and iPhone 15 Pro Max users the ability to record spatial videos. The new feature lets users film in three dimensions and experience their favorite memories and special moments on Apple Vision Pro, the upcoming mixed-reality headset.
In order to create a three-dimensional video, Apple explains that the iPhone uses both the main and ultrawide cameras when recording. This is then saved as a single file within a new album in the Photos app titled “Spatial.” The videos will also sync across devices with iCloud. Spatial videos are captured in 1080p resolution at 30 frames per second.
Spatial video recording can be enabled in Settings by toggling on “Spatial Video for Apple Vision Pro” in the Camera section under Formats. Apple suggests holding the iPhone in landscape orientation for optimal results. ...
The AI Act takes a “risk-based approach” to products or services that use artificial intelligence and focuses on regulating uses of AI rather than the technology. The legislation is designed to protect democracy, the rule of law and fundamental rights like freedom of speech, while still encouraging investment and innovation.
The riskier an AI application is, the stiffer the rules. Those that pose limited risk, such as content recommendation systems or spam filters, would have to follow only light rules such as revealing that they are powered by AI.
High-risk systems, such as medical devices, face tougher requirements like using high-quality data and providing clear information to users.
Some AI uses are banned because they’re deemed to pose an unacceptable risk, like social scoring systems that govern how people behave, some types of predictive policing and emotion recognition systems in school and workplaces. ...
The AI Act won’t take effect until two years after final approval from European lawmakers, expected in a rubber-stamp vote in early 2024. Violations could draw fines of up to 35 million euros ($38 million) or 7% of a company’s global revenue. ...
... Watch for companies with the best blend of prompt applications for storytelling ...
What is undeniable, however, is that the expertise for targeting the appropriate demographic for the content will be critical, and businesses that possess that capability will be the winners. We will also see the rise of even more powerful algorithms that help match content with viewers that rivals even the best of today’s current streaming market.
Watch for entertainment conglomerates that can create the best systems to manage their internal data ...
Watch for stunning startups ...
The worst place to be is in the middle. If you don’t have the corporate size to drive large scale and you don’t have the nimbleness and agility of a startup, the middle is a very, very dangerous place. No one wins in that space, ever.” ...
Additional narratives to note
There were also several discussions around DEI as it pertain to ethics, biases, investors, and founders. ...
Many consistently ask about guardrails to which there is are industry-wide answers for such questions at the moment. Companies that figure out such solutions just may actually become the most valuable yet in this powerful new sector.
Three technology trends besides “pure AI” that I’m tracking include the convergence of physical and digital worlds, the data ecosystem for insights developed with transparency and trust, and the bidirectional feedback loop of science and technology. ...
The bidirectional feedback loop of science and technology enables more rapid innovation to address challenges in the global supply chain, climate change, and medicine. Specifically, I’m tracking advanced materials for energy and exploring the geospatial frontier. For example, the ambitious push for autonomous electric vehicles hinges on the availability of advanced battery technology. Could the solution lie in exploring new materials, nanotechnological, self-healing capabilities, or innovative approaches to energy storage? While we may not have all the answers yet, it's clear that addressing this bottleneck is crucial for supply chains, mobility solutions and climate impact.
... The OEP [ Orlando Economic Partnership ] digital twin project marks the largest of its kind, simulating 800 square miles of the famed region as a 3D render that OEP staff or visitors can interact with via a large-scale screen.
The OEP displays the digital twin at its new downtown Orlando, Florida headquarters, ready for potential regional partners to view the area and related core data points, which help visitors understand a place of business growth, all from one room. ...
For example, in April, Dave Rhodes, the GM and senior VP of Digital Twins at Unity, explained how his firm would be able to integrate related XR technologies such as artificial intelligence (AI), machine learning (ML), simulation, and analytics to improve digital twin workflows at OEP and to match the growing number of use cases the team is receiving.
Unity’s digital technology unlocks data from proprietary formats, making merging anything into one virtual experience more effortless and allowing stakeholders to collaborate, simulate, and make better decisions. ...
Providing a resource for U.S. policymakers, a committee of MIT leaders and scholars has released a set of policy briefs that outlines a framework for the governance of artificial intelligence. The approach includes extending current regulatory and liability approaches in pursuit of a practical way to oversee AI.
The aim of the papers is to help enhance U.S. leadership in the area of artificial intelligence broadly, while limiting harm that could result from the new technologies and encouraging exploration of how AI deployment could be beneficial to society.
The main policy paper, “A Framework for U.S. AI Governance: Creating a Safe and Thriving AI Sector,” suggests AI tools can often be regulated by existing U.S. government entities that already oversee the relevant domains. The recommendations also underscore the importance of identifying the purpose of AI tools, which would enable regulations to fit those applications. ...
“We felt it was important for MIT to get involved in this because we have expertise,” says David Goldston, director of the MIT Washington Office. “MIT is one of the leaders in AI research, one of the places where AI first got started. Since we are among those creating technology that is raising these important issues, we feel an obligation to help address them.” ...
Purpose, intent, and guardrails
The main policy brief outlines how current policy could be extended to cover AI, using existing regulatory agencies and legal liability frameworks where possible. The U.S. has strict licensing laws in the field of medicine, for example. It is already illegal to impersonate a doctor; if AI were to be used to prescribe medicine or make a diagnosis under the guise of being a doctor, it should be clear that would violate the law just as strictly human malfeasance would. ...
While the policy framework involves existing agencies, it includes the addition of some new oversight capacity as well. For one thing, the policy brief calls for advances in auditing of new AI tools, which could move forward along a variety of paths, whether government-initiated, user-driven, or deriving from legal liability proceedings. ...
“AI enables things humans cannot do, such as surveillance or fake news at scale, which may need special consideration beyond what is applicable for humans,” Ozdaglar says. “But our starting point still enables you to think about the risks, and then how that risk gets amplified because of the tools.” ...
As Goldston puts it, the committee releasing these papers is “is not a group that is antitechnology or trying to stifle AI. But it is, nonetheless, a group that is saying AI needs governance and oversight. That’s part of doing this properly. These are people who know this technology, and they’re saying that AI needs oversight.” ...
... The battle between the e/accs and the Effective Altruists is one of many quasi-religious schisms breaking out in San Francisco’s A.I. scene these days, as insiders argue about how quickly the technology is progressing, and what should be done about it.
E/acc prefers the all-gas, no-brakes approach. Its adherents favor open-sourcing A.I. software rather than having it be controlled by big corporations, and unlike Effective Altruists, they don’t see powerful A.I. as something to be feared or guarded against. They believe that A.I.’s benefits far outweigh its harms, and that the right thing to do with such important technology is to get out of the way and let it rip. ...
Initially, I wrote the movement off as a fringe novelty — a bunch of Twitter-addicted techies with persecution complexes turning warmed-over Ayn Rand into edgy memes.
But a few months later, tech luminaries like Marc Andreessen, the co-founder of the venture capital firm Andreessen Horowitz, started showing up in e/acc’s Twitter Spaces, and proclaiming that he, too, believed in effective accelerationism. ...
Critics have pointed to the fact that some of e/acc’s leaders, including Mr. Verdon, seem to actually agree with the Effective Altruists that a rogue A.I. could wipe out humanity, but aren’t bothered by the idea, since superhuman A.I. could represent a logical next step in evolution. ...
