Connect with us


Artificial intelligence, Geoffrey Hinton, neural network, GLOM, vectors, visual perception, human perception, intuition



Hinton face grid

Deep learning set off the latest AI revolution, transforming computer vision and the field as a whole. Hinton believes deep learning should be almost all that’s needed to fully replicate human intelligence.

But despite rapid progress, there are still major challenges. Expose a neural net to an unfamiliar data set or a foreign environment, and it reveals itself to be brittle and inflexible. Self-driving cars and essay-writing language generators impress, but things can go awry. AI visual systems can be easily confused: a coffee mug recognized from the side would be an unknown from above if the system had not been trained on that view; and with the manipulation of a few pixels, a panda can be mistaken for an ostrich, or even a school bus.

GLOM addresses two of the most difficult problems for visual perception systems: understanding a whole scene in terms of objects and their natural parts; and recognizing objects when seen from a new viewpoint.(GLOM’s focus is on vision, but Hinton expects the idea could be applied to language as well.)

An object such as Hinton’s face, for instance, is made up of his lively if dog-tired eyes (too many people asking questions; too little sleep), his mouth and ears, and a prominent nose, all topped by a not-too-untidy tousle of mostly gray. And given his nose, he is easily recognized even on first sight in profile view.

Both of these factors—the part-whole relationship and the viewpoint—are, from Hinton’s perspective, crucial to how humans do vision. “If GLOM ever works,” he says, “it’s going to do perception in a way that’s much more human-like than current neural nets.”

Grouping parts into wholes, however, can be a hard problem for computers, since parts are sometimes ambiguous. A circle could be an eye, or a doughnut, or a wheel. As Hinton explains it, the first generation of AI vision systems tried to recognize objects by relying mostly on the geometry of the part-whole-relationship—the spatial orientation among the parts and between the parts and the whole. The second generation instead relied mostly on deep learning—letting the neural net train on large amounts of data. With GLOM, Hinton combines the best aspects of both approaches.

“There’s a certain intellectual humility that I like about it,” says Gary Marcus, founder and CEO of Robust.AI and a well-known critic of the heavy reliance on deep learning. Marcus admires Hinton’s willingness to challenge something that brought him fame, to admit it’s not quite working. “It’s brave,” he says. “And it’s a great corrective to say, ‘I’m trying to think outside the box.’”

The GLOM architecture

In crafting GLOM, Hinton tried to model some of the mental shortcuts—intuitive strategies, or heuristics—that people use in making sense of the world. “GLOM, and indeed much of Geoff’s work, is about looking at heuristics that people seem to have, building neural nets that could themselves have those heuristics, and then showing that the nets do better at vision as a result,” says Nick Frosst, a computer scientist at a language startup in Toronto who worked with Hinton at Google Brain.

With visual perception, one strategy is to parse parts of an object—such as different facial features—and thereby understand the whole. If you see a certain nose, you might recognize it as part of Hinton’s face; it’s a part-whole hierarchy. To build a better vision system, Hinton says, “I have a strong intuition that we need to use part-whole hierarchies.” Human brains understand this part-whole composition by creating what’s called a “parse tree”—a branching diagram demonstrating the hierarchical relationship between the whole, its parts and subparts. The face itself is at the top of the tree, and the component eyes, nose, ears, and mouth form the branches below.

One of Hinton’s main goals with GLOM is to replicate the parse tree in a neural net—this would distinguish it from neural nets that came before. For technical reasons, it’s hard to do. “It’s difficult because each individual image would be parsed by a person into a unique parse tree, so we would want a neural net to do the same,” says Frosst. “It’s hard to get something with a static architecture—a neural net—to take on a new structure—a parse tree—for each new image it sees.” Hinton has made various attempts. GLOM is a major revision of his previous attempt in 2017, combined with other related advances in the field.

“I’m part of a nose!”

GLOM vector


A generalized way of thinking about the GLOM architecture is as follows: The image of interest (say, a photograph of Hinton’s face) is divided into a grid. Each region of the grid is a “location” on the image—one location might contain the iris of an eye, while another might contain the tip of his nose. For each location in the net there are about five layers, or levels. And level by level, the system makes a prediction, with a vector representing the content or information. At a level near the bottom, the vector representing the tip-of-the-nose location might predict: “I’m part of a nose!” And at the next level up, in building a more coherent representation of what it’s seeing, the vector might predict: “I’m part of a face at side-angle view!”

But then the question is, do neighboring vectors at the same level agree? When in agreement, vectors point in the same direction, toward the same conclusion: “Yes, we both belong to the same nose.” Or further up the parse tree. “Yes, we both belong to the same face.”

Seeking consensus about the nature of an object—about what precisely the object is, ultimately—GLOM’s vectors iteratively, location-by-location and layer-upon-layer, average with neighbouring vectors beside, as well as predicted vectors from levels above and below.

However, the net doesn’t “willy-nilly average” with just anything nearby, says Hinton. It averages selectively, with neighboring predictions that display similarities. “This is kind of well-known in America, this is called an echo chamber,” he says. “What you do is you only accept opinions from people who already agree with you; and then what happens is that you get an echo chamber where a whole bunch of people have exactly the same opinion. GLOM actually uses that in a constructive way.” The analogous phenomenon in Hinton’s system is those “islands of agreement.”

“Geoff is a highly unusual thinker…”

Sue Becker

“Imagine a bunch of people in a room, shouting slight variations of the same idea,” says Frosst—or imagine those people as vectors pointing in slight variations of the same direction. “They would, after a while, converge on the one idea, and they would all feel it stronger, because they had it confirmed by the other people around them.” That’s how GLOM’s vectors reinforce and amplify their collective predictions about an image.

GLOM uses these islands of agreeing vectors to accomplish the trick of representing a parse tree in a neural net. Whereas some recent neural nets use agreement among vectors for activation, GLOM uses agreement for representation—building up representations of things within the net. For instance, when several vectors agree that they all represent part of the nose, their small cluster of agreement collectively represents the nose in the net’s parse tree for the face. Another smallish cluster of agreeing vectors might represent the mouth in the parse tree; and the big cluster at the top of the tree would represent the emergent conclusion that the image as a whole is Hinton’s face. “The way the parse tree is represented here,” Hinton explains, “is that at the object level you have a big island; the parts of the object are smaller islands; the subparts are even smaller islands, and so on.”

Figure 2 from Hinton’s GLOM paper. The islands of identical vectors (arrows of the same color) at the various levels represent a parse tree.


According to Hinton’s long-time friend and collaborator Yoshua Bengio, a computer scientist at the University of Montreal, if GLOM manages to solve the engineering challenge of representing a parse tree in a neural net, it would be a feat—it would be important for making neural nets work properly. “Geoff has produced amazingly powerful intuitions many times in his career, many of which have proven right,” Bengio says. “Hence, I pay attention to them, especially when he feels as strongly about them as he does about GLOM.”

The strength of Hinton’s conviction is rooted not only in the echo chamber analogy, but also in mathematical and biological analogies that inspired and justified some of the design decisions in GLOM’s novel engineering.

“Geoff is a highly unusual thinker in that he is able to draw upon complex mathematical concepts and integrate them with biological constraints to develop theories,” says Sue Becker, a former student of Hinton’s, now a computational cognitive neuroscientist at McMaster University. “Researchers who are more narrowly focused on either the mathematical theory or the neurobiology are much less likely to solve the infinitely compelling puzzle of how both machines and humans might learn and think.”

Turning philosophy into engineering

So far, Hinton’s new idea has been well received, especially in some of the world’s greatest echo chambers. “On Twitter, I got a lot of likes,” he says. And a YouTube tutorial laid claim to the term “MeGLOMania.”

Hinton is the first to admit that at present GLOM is little more than philosophical musing (he spent a year as a philosophy undergrad before switching to experimental psychology). “If an idea sounds good in philosophy, it is good,” he says. “How would you ever have a philosophical idea that just sounds like rubbish, but actually turns out to be true? That wouldn’t pass as a philosophical idea.” Science, by comparison, is “full of things that sound like complete rubbish” but turn out to work remarkably well—for example, neural nets, he says.

GLOM is designed to sound philosophically plausible. But will it work?


The hunter-gatherer groups at the heart of a microbiome gold rush



The hunter-gatherer groups at the heart of a microbiome gold rush

The first step to finding out is to catalogue what microbes we might have lost. To get as close to ancient microbiomes as possible, microbiologists have begun studying multiple Indigenous groups. Two have received the most attention: the Yanomami of the Amazon rainforest and the Hadza, in northern Tanzania. 

Researchers have made some startling discoveries already. A study by Sonnenburg and his colleagues, published in July, found that the gut microbiomes of the Hadza appear to include bugs that aren’t seen elsewhere—around 20% of the microbe genomes identified had not been recorded in a global catalogue of over 200,000 such genomes. The researchers found 8.4 million protein families in the guts of the 167 Hadza people they studied. Over half of them had not previously been identified in the human gut.

Plenty of other studies published in the last decade or so have helped build a picture of how the diets and lifestyles of hunter-gatherer societies influence the microbiome, and scientists have speculated on what this means for those living in more industrialized societies. But these revelations have come at a price.

A changing way of life

The Hadza people hunt wild animals and forage for fruit and honey. “We still live the ancient way of life, with arrows and old knives,” says Mangola, who works with the Olanakwe Community Fund to support education and economic projects for the Hadza. Hunters seek out food in the bush, which might include baboons, vervet monkeys, guinea fowl, kudu, porcupines, or dik-dik. Gatherers collect fruits, vegetables, and honey.

Mangola, who has met with multiple scientists over the years and participated in many research projects, has witnessed firsthand the impact of such research on his community. Much of it has been positive. But not all researchers act thoughtfully and ethically, he says, and some have exploited or harmed the community.

One enduring problem, says Mangola, is that scientists have tended to come and study the Hadza without properly explaining their research or their results. They arrive from Europe or the US, accompanied by guides, and collect feces, blood, hair, and other biological samples. Often, the people giving up these samples don’t know what they will be used for, says Mangola. Scientists get their results and publish them without returning to share them. “You tell the world [what you’ve discovered]—why can’t you come back to Tanzania to tell the Hadza?” asks Mangola. “It would bring meaning and excitement to the community,” he says.

Some scientists have talked about the Hadza as if they were living fossils, says Alyssa Crittenden, a nutritional anthropologist and biologist at the University of Nevada in Las Vegas, who has been studying and working with the Hadza for the last two decades.

The Hadza have been described as being “locked in time,” she adds, but characterizations like that don’t reflect reality. She has made many trips to Tanzania and seen for herself how life has changed. Tourists flock to the region. Roads have been built. Charities have helped the Hadza secure land rights. Mangola went abroad for his education: he has a law degree and a master’s from the Indigenous Peoples Law and Policy program at the University of Arizona.

Continue Reading


The Download: a microbiome gold rush, and Eric Schmidt’s election misinformation plan



The Download: a microbiome gold rush, and Eric Schmidt’s election misinformation plan

Over the last couple of decades, scientists have come to realize just how important the microbes that crawl all over us are to our health. But some believe our microbiomes are in crisis—casualties of an increasingly sanitized way of life. Disturbances in the collections of microbes we host have been associated with a whole host of diseases, ranging from arthritis to Alzheimer’s.

Some might not be completely gone, though. Scientists believe many might still be hiding inside the intestines of people who don’t live in the polluted, processed environment that most of the rest of us share. They’ve been studying the feces of people like the Yanomami, an Indigenous group in the Amazon, who appear to still have some of the microbes that other people have lost. 

But there is a major catch: we don’t know whether those in hunter-gatherer societies really do have “healthier” microbiomes—and if they do, whether the benefits could be shared with others. At the same time, members of the communities being studied are concerned about the risk of what’s called biopiracy—taking natural resources from poorer countries for the benefit of wealthier ones. Read the full story.

—Jessica Hamzelou

Eric Schmidt has a 6-point plan for fighting election misinformation

—by Eric Schmidt, formerly the CEO of Google, and current cofounder of philanthropic initiative Schmidt Futures

The coming year will be one of seismic political shifts. Over 4 billion people will head to the polls in countries including the United States, Taiwan, India, and Indonesia, making 2024 the biggest election year in history.

Continue Reading


Navigating a shifting customer-engagement landscape with generative AI



Navigating a shifting customer-engagement landscape with generative AI

A strategic imperative

Generative AI’s ability to harness customer data in a highly sophisticated manner means enterprises are accelerating plans to invest in and leverage the technology’s capabilities. In a study titled “The Future of Enterprise Data & AI,” Corinium Intelligence and WNS Triange surveyed 100 global C-suite leaders and decision-makers specializing in AI, analytics, and data. Seventy-six percent of the respondents said that their organizations are already using or planning to use generative AI.

According to McKinsey, while generative AI will affect most business functions, “four of them will likely account for 75% of the total annual value it can deliver.” Among these are marketing and sales and customer operations. Yet, despite the technology’s benefits, many leaders are unsure about the right approach to take and mindful of the risks associated with large investments.

Mapping out a generative AI pathway

One of the first challenges organizations need to overcome is senior leadership alignment. “You need the necessary strategy; you need the ability to have the necessary buy-in of people,” says Ayer. “You need to make sure that you’ve got the right use case and business case for each one of them.” In other words, a clearly defined roadmap and precise business objectives are as crucial as understanding whether a process is amenable to the use of generative AI.

The implementation of a generative AI strategy can take time. According to Ayer, business leaders should maintain a realistic perspective on the duration required for formulating a strategy, conduct necessary training across various teams and functions, and identify the areas of value addition. And for any generative AI deployment to work seamlessly, the right data ecosystems must be in place.

Ayer cites WNS Triange’s collaboration with an insurer to create a claims process by leveraging generative AI. Thanks to the new technology, the insurer can immediately assess the severity of a vehicle’s damage from an accident and make a claims recommendation based on the unstructured data provided by the client. “Because this can be immediately assessed by a surveyor and they can reach a recommendation quickly, this instantly improves the insurer’s ability to satisfy their policyholders and reduce the claims processing time,” Ayer explains.

All that, however, would not be possible without data on past claims history, repair costs, transaction data, and other necessary data sets to extract clear value from generative AI analysis. “Be very clear about data sufficiency. Don’t jump into a program where eventually you realize you don’t have the necessary data,” Ayer says.

The benefits of third-party experience

Enterprises are increasingly aware that they must embrace generative AI, but knowing where to begin is another thing. “You start off wanting to make sure you don’t repeat mistakes other people have made,” says Ayer. An external provider can help organizations avoid those mistakes and leverage best practices and frameworks for testing and defining explainability and benchmarks for return on investment (ROI).

Using pre-built solutions by external partners can expedite time to market and increase a generative AI program’s value. These solutions can harness pre-built industry-specific generative AI platforms to accelerate deployment. “Generative AI programs can be extremely complicated,” Ayer points out. “There are a lot of infrastructure requirements, touch points with customers, and internal regulations. Organizations will also have to consider using pre-built solutions to accelerate speed to value. Third-party service providers bring the expertise of having an integrated approach to all these elements.”

Continue Reading

Copyright © 2021 Seminole Press.