There are enormous opportunities in this field, and we are only at the beginning of this creative revolution.
Innumerous AI systems are advancing in their ability to “create,” in some cases muddying the surrounding legal waters. I have selected here some exciting projects that are starting to explore this trend in the AI landscape:
Google revealed it was training its AI to design computer chips with optimized chip placements in less than 6 hours. In comparison, human counterparts might take weeks to create similar or inferior chips.
LegalMation, which has an AI platform that drafts legal paperwork, updated its platform to prepare answers and discovery documents in response to COVID-19-related causes of action. Last year, its competitor Casetext launched an AI-powered automated legal brief-drafting platform called Compose and raised $8.2M from investors.
Audio deepfakes, many mimicking famous personalities, and music have advanced in quality and grown in popularity.
YouTube received a copyright claim regarding an audio deepfake seeking a content takedown.
Entertainment agency Roc Nation said regarding two-deep faked audio clips of Jay-Z reciting Shakespeare and singing Billy Joel: “This content unlawfully uses an AI to impersonate our client’s voice.” YouTube initially took down the videos but later put them back up due to the claim’s questionable legal validity.
Known for its text-generating GPT-2 AI system, OpenAI released a music-generating neural net called Jukebox to generate music samples in various genres and styles, with artificially generated lyrics co-written by the system and AI researchers.
Aimi launched an AI-powered app in a similar vein that can turn real tracks from artists into adaptive electronic music to meet listeners’ needs. Its competitors include Endel (which has a deal with Warner Music), Weav Run, and Mubert.
Startup Codota’s AI-powered code completion is helping developers write 25% less code. Other code completion tools include Kite (which has 250K+ users), Deep TabNine, and IntelliSense.
Historically, automated code generation has had a poor reputation. Still, there are growing signals that AI is productively employed across the software development process, including AI-assisted programming and supporting low– and no-code development.
Electronic Arts announced it was experimenting with AI to automate lifelike animated characters (such as a controllable soccer player) and accelerate its video-game development process. Separately, researchers from the University of Toronto and MIT collaborated with Nvidia and announced an AI engine that could recreate a new Pac-Man version without access to the original code.
Microsoft just announced it was laying off dozens of journalists and editorial staff at Microsoft News and MSN in favor of AI-powered content curation. About 50 teams in the US and 27 in the UK (it had 800 editorial staff in late 2018) have been reportedly affected.
The past week saw two AI drug discovery startups (Intro, Exscientia) raise $200M+ in funding.
Another drug development firm Syntekabio announced they had used their AI platform to screen 3,000 drugs for potential COVID-19 therapeutics.
These are just the latest developments among the 230+ startups and big pharma using AI to discover new drugs from antibiotics to antibody therapies, in some cases far faster and for far less than the typical 4.5 years in development time and $2.6B in cost.
Many media outlets — including The Washington Post, Bloomberg, Reuters, Associated Press, and The Guardian — use AI systems to automate content generation to varying degrees.
The quality of some forms of AI-powered text generation — such as OpenAI’s GPT-2 — has advanced to the point where there are concerns about misuse.
As AI advances and becomes more responsible for concept and content generation, the extent to which its output can be protected as the intellectual property will have vast implications. But I will cover this in my next article.
The possibilities in this field are endless, and we are only at the beginning of this creative revolution.
The more we understand Artificial Intelligence as an ally and not as an enemy, the more creative the work generated by professionals will be, taking advantage of this new technological wave instead of claiming it exists.
If you want to read more about creative AI, deep learning, and other topics here you have some other articles:
Also, I’ve just published a new ebook on Amazon, and I’m already working on publishing some other new ones across this year… let’s keep in touch, and let’s do it together.