|
|
Toward a US-Wide Research Cloud? |
Several U.S. companies (Google, Amazon, IBM) and universities (Stanford, Carnegie-Mellon, Ohio State) are joining an effort
launched by several lawmakers on June 4 to create a National Research Cloud. This would make resources from large cloud providers available to academic researchers. The project bears some similarity to the ARPAnet (which, after several iterations, became the Internet) of 50 years ago. The lawmakers' press release (see above link) also explicitly invokes the risk of China's dominance in AI, thus recalling the quasi-panic about Japanese dominance of chip manufacturing that led to the creation of the SEMATECH research center in 1988. Thus, true to form, we are going to make this a matter of national pride and go at it alone instead of collaborating with other countries.
Just like Stanford's Vint Cerf played a key role in the architecture of the ARPAnet, two Stanford people are the key proponents of the new initiative: Prof. and former Provost John Etchemendy and Prof. Fei-Fei Li, co-founders of the Stanford Institute for Human-Centered Artificial Intelligence.
Of course, since this is the government, this initiative has to start with forming a task force, which will draw up a plan, which will... you get the idea. While our tax dollars should not be wasted, perhaps some of these people should read the
Agile Manifesto.
|
|
Virtual Computer Science Conferences |
One of the few positive effects of the current COVID-related lockdowns is that it is much easier to attend conferences that have moved to a virtual format. Registration fees have been considerably reduced or eliminated in many cases, and there are no travel costs. ACM just sent the following announcements of online conferences:
- GECCO-2020 (Genetic and Evolutionary Computing Conference), July 8-12. When robots design other robots, software evolves through natural selection, and more.
- EC-2020 (Conference on Economics and Computation), July 13-17. The economics of networks, systems and computation.
- DAC 2020 (Design Automation Conference), July 20-24. Computer-aided design for everything, including new computer forms (quantum, neuromorphic, ...).
- L@S 2020 (Learning at Scale), August 12-14. Improving online learning environments and personalizing the learner's experience.
As for SIGGRAPH Asia (officially the "Conference and Exhibition on Computer Graphics and Interactive Techniques in Asia"), it is still planned as an in-person event in Daegu, South Korea, on November 17-20. But as Dr. Fauci said, "we don't decide for the virus, the virus decides for us." It is safe to assume that the organizers are already making contingency plans in case packing a convention center is still a bad idea then. |
|
3D Printing for Formula 1 Car Parts
|
In this short video, Renault engineers explain (in very British, not French, accents) the role that additive manufacturing, using 3D printing equipment from 3D Systems, a South Carolina company, plays in creating the custom parts needed for each year's new model of Grand Prix race cars.
Thanks to Joel Rosenberg, Expert in Smart Technology for the European Commission (among other titles) for sharing this.
|
|
KMWorld Report on Content and Knowledge Management
|
Earlier this year, KMWorld -- the organization behind the annual conference of the same name -- published a report entitled "Looking to the Future: 2020 Insight." We read it and digested it for you.
On the negative side, there are lots of platitudes or head-scratching pronouncements (here's a gem: "over the course of the past decade, digitalization has created an undeniable shift to bring businesses to where they are now." Duh.) The conclusion section is also an accumulation of obvious statements. This said, we extracted a few useful points:
- Suppliers and customers appear to focus on customer experience -- specifically, arming call centers with the knowledge to answer customer questions. Much less attention is given to the internal sharing of knowledge in non-customer facing functions (engineering, manufacturing, maintenance, etc.) We think this is a mistake, especially considering how many people whose knowledge resides only in their heads are being laid off during crises such as the current one.
- There's a bit of a "flight forward" attitude -- people haven't solved their knowledge capture and sharing yet, but they're already talking about adding AI to the mix, clearly viewing it as a magic bullet.
- The market is growing quickly. "According to Fortune Business Insights, the global enterprise content management market size, which includes content workflow, document management, imaging and capturing, web content management, record management, mobile content management, digital asset management, and case management, was valued at $15.33 billion in 2018 and is projected to reach $43.16 billion by 2026 for a compound annual growth rate of 14%" (you will have noticed the use of completely useless precision in the target number, one of our pet peeves).
- Data protection issues are well known by now, but the paper contains useful and specific estimates of the costs of compliance with regulations such as Europe's GDPR and California's CCPA -- and the cost of remedying data breaches when they occur.
- The paper mentions knowledge graphs as an important technology to organize information and extend taxonomies and ontologies, but gives very little detail.
|
|
Seen Recently... |
"We shouldn't presume that a group of experts somehow knows what's best."
-- U.S. Senator Rand Paul, in a widely mocked comment made during a
legislative hearing. The more charitable responses on the Internet
boiled down to "please look up the definition of the word 'expert'."
|
|
|
|