We're excited to share that Cognizance Technologies has been awarded an NIH NIAID NEAT BPA Order for AV Events Support Services.
Another step forward in supporting federal health missions.
#HealthIT#GovCon#NIH
A study was conducted not too long ago suggesting that we (humans) are ok to let algorithms make decisions for us (not all but many of them). The time is right then for NIST to start working on a Risk Management Framework for AI -
https://t.co/4xW5zbORXY
"...degree to which human beings can understand the cause of a decision" or "...the degree to which a machine's output can be consistently predicted" - great write-up about understanding and the need for interpretability.
https://t.co/vG9UbFfkzf
Of course, we all know about and of Paul Gauguin, but have you seen GauGAN2? As in, NVIDIA's "paint me a picture" AI model? Check it out...
https://t.co/0vHKF0hkOD
"The strategy identifies VA as a unique incubator for AI projects because it serves as the largest repository of health care data in the country and operates an integrated health care system of medical centers across the country."
https://t.co/4o58KlH0eI
What's the most important building block of data and AI applications? Data, of course. What's most important about data? Security. Securely leveraging data is the key to building the next generation of applications.
https://t.co/AjrSucJSis
Using open source or publicly available assets has always been "something to check and make sure is ok to do". Using publicly available datasets is no different. Just because it's public, doesn't mean you can use it in your software and make a profit.
https://t.co/YDuurHLFvF
NIST's Privacy Engineering Program has series of posts on Differential Privacy. They're up to 11 articles now - each better than the one before it. Check it out - differential privacy is very helpful.
https://t.co/N24TxtOucs
Staying with the quality theme, here is another article (this one from TheNewStack) talking about using energy usage metrics as quality determinants. A program written in 27 programming languages was analyzed and the question of quality was reviewed.
https://t.co/2HNAx5R9L7
Here is a question...if the code is written by AI - for example by GitHub Copilot - does it matter if it's good or bad? Would you care?
https://t.co/oTGlIcvyIC
AWS on your laptop - that's what localstack provides - a very handy and useful tool. If you do cloud (and if you don't, you need to ping us) and haven't looked at this yet, you'll be pleasantly surprised.
https://t.co/DWKeZrVgkO
Unfinished 10th...finished?
"But the phrases he was listening to weren’t composed by Beethoven. They were created by artificial intelligence—a computer simulation of Beethoven’s creative process."
https://t.co/70tetCmtWo
You know how developers have a definition of "done"? When it comes to data, it is through the quality program that you assess how good that data is.
And so the question is, what does good actually look like? well...here is a start -
https://t.co/PMhjfgUvO0
An analysis of developments in AI that lead us to have debates on bots taking over. This one is from across the pond.
State of AI Report 2021
https://t.co/ixFaeUEAiV
"...we can’t simply deploy shiny new systems and call it modernization. We have to modernize the workforce, to include accessibility right up front in the design phase"
https://t.co/NEnIC0wlE2
AI is becoming very much commonplace. And the black box mentality surrounding it is no longer acceptable. In this article Linked-in Engineering is presenting their method for making AI transparent and explainable.
https://t.co/hWsvv8viGw
So...could AI help with writing contracts? DoD is offering 50K to develop a prototype. It would be interesting to see what the submitted ideas are.
https://t.co/aACg2jtgvD
NIST is holding a very important workshop starting on Oct. 19 for working towards creating an AI Risk Mgmt. Framework. It's a 3-day event during which you can help develop an AI Framework to manage risks to individuals, organizations, and society.
https://t.co/v8xKzYoH9L