A major approach for defending against adversarial attacks aims at controlling only image classifiers to be more resilient, and it does not care about visual objects, such as pandas and cars, in images. This means that visual objects themselves cannot take any defensive actions, and they are still vulnerable to adversarial attacks. In contrast, letters are artificial symbols, and we can freely control their appearance unless losing their readability. In other words, we can make the letters more defensive to the attacks. This paper poses three research questions related to the adversarial vulnerability of letter images: (1) How defensive are the letters against adversarial attacks? (2) Can we estimate how defensive a given letter image is before attacks? (3) Can we control the letter images to be more defensive against adversarial attacks? For answering the first and second questions, we measure the defensibility of letters by employing Iterative Fast Gradient Sign Method (I-FGSM) and then build a deep regression model for estimating the defensibility of each letter image. We also propose a two-step method based on a generative adversarial network (GAN) for generating character images with higher defensibility, which solves the third research question.

Identifying healthcare fraud, waste and abuse is a highly evolved practice that is best done with a marriage of artificial intelligence (AI) and human intelligence (HI) capabilities. As the losses attributed to fraud continues to grow, unfortunately, we all share the responsibility of paying for it. The National Health Care Anti-Fraud Association (NHCAA) estimates that the financial losses due to health care fraud are in the tens of billions of dollars each year.1 The payment integrity review process of analyzing a healthcare claim can be strengthened by implementing a hybrid approach of both HI and AI. However, it's important to understand the benefits and limitations of each to avoid pitfalls that can arise. On July 20, 2022, the Department of Justice announced criminal charges against 36 defendants in 13 federal districts across the United States for more than $1.2 billion in alleged fraudulent telemedicine, cardiovascular and cancer genetic testing, and durable medical equipment (DME) schemes.

The funding environment for AI startups remains robust--and some of the rounds have been substantial. Just recently Dataiku, which operates a machine learning platform, announced a $100 million Series D investment. AI truly represents a transformation in the tech world. "AI is making software much more dynamic and improves as it understands user behavior," said Gordon Ritter, who is the founder and General Partner at Emergence. OK then, what are VCs looking for when evaluating an AI deal?

The funding environment for AI startups remains robust--and some of the rounds have been substantial. Just recently Dataiku, which operates a machine learning platform, announced a $100 million Series D investment. AI truly represents a transformation in the tech world. "AI is making software much more dynamic and improves as it understands user behavior," said Gordon Ritter, who is the founder and General Partner at Emergence. OK then, what are VCs looking for when evaluating an AI deal?

Since the term was coined in 1956, "Artificial Intelligence" has endured a lifetime of misunderstanding. The root of the problem lies in the interpretation of the word "intelligence." In the words of legendary computer scientist Edsger Dijkstra: "The question of whether a computer can think is no more interesting than the question of whether a submarine can swim." If I were to ask you who is a better swimmer, would you choose Michael Phelps or the USS Illinois? Swimming is a human activity, certainly not something a submarine does.

Building a B2B AI startup is hard enough between struggling to obtain training data and fighting with major tech companies to secure talent. Building a B2B AI startup held to the well-established software-as-a-service (SaaS) metrics is even harder. While many AI businesses deliver value via software monetized by a recurring subscription like their SaaS counterparts, the similarities between the two types of businesses end there. SaaS products built without data and AI offer generalized solutions to their customers. AI businesses more closely resemble a services business or consultancies because they provide solutions that become tailored to that customer's specific needs.

Summary: The fourth and final AI strategy we'll review is Systems of Intelligence (SOI). This is getting nearly as much attention as the Vertical strategy we previously reviewed. It's appealing because it seems to offer the financial advantages of a Horizontal strategy but its ability to create a defensible moat requires some fine tuning. In the last several articles we've been looking at different strategies for successful AI companies. We described Data Dominance, and the Vertical and Horizontal strategies.

One thing that you learn quickly when you enter VC is that investments are about finding moats. I could not find a better way to explain it than paraphrasing Gil Dibner in this post. Why are moats a proxy for the ability to generate profits? Simply, because moats increase a firm's bargaining power with both their suppliers and their customers, helping the firm increase prices and reduce costs in order to generate higher profits. The result of this simple reasoning is that VCs look for companies creating moats.

A few weeks ago I published a post in which I outline a framework to think about ways to create moats for AI startups. It triggered interesting conversations with several AI practitioners (founders, product managers, VCs) over the past few days. I think these debates bring interesting complements to my initial framework, so I eventually decided to publish 3 of them here. I hope these can be of interest for those who read my initial post. In case you've missed it, here it is.

ISC (RotM) Artificial intelligence is about to transform society in the same way electricity did 100 years ago, but researchers are nowhere near producing the sort of self-aware sociopathic systems beloved of sci-fi writers. At least that's what Andrew Ng, Silicon Valley-based chief scientist at Chinese Web giant Baidu, when he kicked off the International Supercomputing Conference, by sketching the progress of neural networks, or deep learning platforms over the last decade. Ng said that in 2007, researchers were working on the CPU level, and were making networks with one million connections. As technology has progressed through the use of GPUs, and onto the cloud, and into the realms of HPC technology, networks were being constructed with 100 million connections. At the same time, he said, researchers were able to use much larger data sets. Whereas academic research projects on speech recognition had worked with data sets of 2000 hours of speech, Baidu's own speech recognition project was using 40,000 hours, he said, resulting in something close to a game-changing 99 per cent accuracy.