How can machine learning answer the centuries-old question of who we really are?
John Bracaglia's Profile
John Bracaglia
Submitted
Activity Feed
Really nice article, thanks for sharing it! Given your description, I imagine a Thingiverse-type site (https://www.thingiverse.com), but stocked with only military weapons and schematics. That would be a nice system to gain access to… While cyber-security is an issue, I’m also curious about the trickle-down effect of the ability to print weapons in civilian society. How can we control access to weapons and dangerous technologies if anyone can print a gun in their own home? Even if this is illegal, is it enforceable?
Super nice article, thanks for sharing it! I agree with your point on the narrowness of their R&D spend – perhaps they should begin by targeting other types of regenerative tissues (e.g. joints, ligaments) that improve quality of life rather than offer vital support. 23% of adults in the US have arthritis, and direct medical costs are in the scale of $140B. Focusing here could greatly reduce their regulatory risk and expand their market base. It could also give them a runway to advance research in more vital tissue applications that they may pursue in the future.
Really nice article, thanks for sharing! This seems to be a legitimate use case for blockchain, which I’m happy to read! I’m curious about the regulatory risk associated with cross-border settlements, especially regarding anti-money laundering rules. AML rules are a primary reason for the length/ scope of validation between foreign banks. If I am able to send money instantaneously across borders, how can I be assured that banks will still meet these requirements, especially given that users of crypto exchanges are often de-identified?
Really great article, thanks for sharing! As an avid user of SoundCloud (and as a paying member), I feel the pain… I’d argue that SoundCloud’s biggest issue is the lack of clear product delineation – right now, they’re charging on both the supply side (artists pay to increase upload limits), and demand side (users pay to access more songs). But neither is a compelling option that scales with their user base growth. The upload limits are too high for most artists to need to pay to use the platform, and SoundCloud Go is essentially useless to consumers because the ‘exclusive’ tracks on the site can be found in a boot-legged version elsewhere… on their own site! A better model might be to create a marketplace for direct purchasing of music, similar to that of BandCamp. I’m rooting for them!
Super nice post, thanks for sharing this! I very much agree with the need delineated at the top – the recent surge of mass scale data leaks (Equifax), fake accounts (Twitter), and privacy-centric regulation (GDPR) have illustrated a real issue with online authentication. There is also a need for KYC (‘Know Your Customer’) in crypto exchanges to reduce fraud. But… if Jumio is creating a centralized database of user profiles and government/ biometric documents, is there a material risk of being hacked? I’d argue that this type of data storage will need to be decentralized, or else they put themselves up as a very large target as their user base continues to grow.
Super nice article, thanks for sharing! I loved your breakout of the digital marketing stack and agree that AI has started to affect many of its levels. I unfortunately also agree that creativity won’t be safe for long… Convolutional neural networks are already trained on millions of examples of a particular type of data (say, synthesizer sounds, or… advertising copy) and learn parameters that are particularly useful in that class of data. They then can generate examples on their own by initializing with random weights and optimizing to create output that look like the training data. Here’s a few examples of different creative coding initiatives 🙂 : https://experiments.withgoogle.com/collection/ai