Hackers are rarely caught. Twitter’s were because they’re American.
The easiest way for a cybercriminal to get caught is to live and operate in America.
Essays on artificial intelligence, machine learning, cybercrime, and the tradeoffs that come with both. Written 2019–2020.
The easiest way for a cybercriminal to get caught is to live and operate in America.
Artificial intelligence and the law are both failing to stop cybercrime today, separately. Putting them together will be key to solving the problem.
AI is the most successful approach yet developed to stop cybercrime. It can’t stop the abuse completely, but it can drive up their costs so that most abusers go out of business.
Legal approaches have generally been ineffective at fighting spam to date, but there are lessons we can learn from this including promising future pathways.
Learning seems easy for humans because we start with rich data more advanced than anything available to computers. It's much harder for computers to learn.
Humans do an enormous amount of work to prepare the data used in AI systems, like a spam filter. Machines need to be provided just the right data in order to learn.
When humans learn, they start with rich, structured data about the world. We perceive objects, and this is the input to consciousness--where we learn.
Artificial intelligence is being integrated into every aspects of our lives. It's the perfect time to learn about this area and prepare for the future.
AI can't swab a patient's nose, but it can make tests available in record time and help find a vaccine. AI has a key supporting role in fighting COVID-19.
...governments need to recognize net states as geopolitical entities... [They] are now actively working in defense, diplomacy, public infrastructure, and citizen services
...the machine can’t learn anything beyond what a human painstakingly teaches it. The algorithm learns by recognizing patterns in the data provided by humans, with that data containing all the right answers.
Cybercrime is a massive, growing problem. AI is the purported solution, but it's not enough. Even security companies won't cut it. Government help is needed
Humans are self-driven and learn from their senses, forming concepts and propositions, with logic as an auditor. Computers learn very differently.
Big tech companies like Google, Facebook, Amazon, and others all have liberal data collection practices. Most of the popular discussions of this are major concerns about privacy and concern about these companies for using personal data for ad targeting. I want to talk about how collecting personal data is critical for keeping you safe online. […]
Humans look at data for signs of bias. Identifying bias—and thus remediating it—is inextricably linked to activities that pose privacy concerns.
Google helps people find niche content [...] Far from stifling the distribution of free content, Google is enabling it on an unprecedented scale.
The world is very structured. Machine learning is not able to take full advantage of little of this in theory, but it does in practice as it picks up on patterns in the data that reflect this real-world structure.
Many concerns are based on misconceptions about artificial intelligence. I'm working to make a deeper understanding of AI's capabilities more accessible.