Lorem ipsum dolor sit amet, consectetur adicing elit ut ullamcorper. leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet. Leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet.

  /  Project   /  Blog: Tech & Telecom news — May 9, 2019

Blog: Tech & Telecom news — May 9, 2019



Disney beat analysts’ expectations for its 1Q19 results, announced yesterday, but losses keep accumulating in the company’s “Direct-to-Consumer” segment, on increased content production expenses for their streaming services (ESPN+ and the forthcoming Disney+). This division already brings almost $1bn/Q (+15% yoy) (Story)


One of the 4 platforms for growth that S Nadella presented this week at Microsoft’s Build event was the “Power Platform”, a new set of tools oriented to build and manage an ecosystem of apps for process automation, that clients would get in a bundle with Dynamics 365 (the company’s suite for CRM and ERP) (Story)



The “neutral 5G wholesale network” scenario that was discussed in the US earlier this year, after comments by Trump’s staff, looks increasingly improbable. An FCC commissioner opposed to such model yesterday, claiming that the proposal “borders on the preposterous” and is practically not feasible, for several reasons (Story)

The US keeps pressing its allies to ban Huawei from new 5G networks, and at London yesterday the US Secretary of State said that current decisions to allow Huawei equipment in some network segments could allow China to “divide Western alliances” and “impede US ability to share certain information” (Story)


Artificial Intelligence

Chinese government’s plans to use facial recognition technologies to implement a massive surveillance system is being reflected in the (huge) valuations of two local startups specialized in this: SenseTime and Megvii. The latter announced yesterday that they have just raised $750m on a valuation of more than $4bn (Story)

A major AI research conference this week focused on 4 deep learning challenges: fairness (eliminating algorithmic bias), security (fighting vs. attacks from “malevolent” AI systems), generalizability (learning from one data set for multiple apps) and causality (finding causal relations in data, instead of correlations) (Story)


As expected, in a hearing yesterday at US Congress, there was debate on how to address the (Facebook) data privacy issue. Democrats want Federal regulators to have more authority to set rules and impose fines, but Republicans favor a less interventionist approach, focused on adapting rules as industry practices change (Story)

An FT piece remarks that data is not “the new oil” but just a way to code (some) real world facts, and that indeed many incidences of “weaponized information” are non-digital. Author also claims that valuing data as a commodity would be subject to hyperinflation, as people could code data about themselves endlessly (Story)

Whatever you may think about that, Google has clearly prioritized data and privacy in messages at their event this week. Analyst B Thompson (Stratechery) thinks the company is trying to earn the right to collect and analyze data, by showing it can be a tool to create more useful products, rather than a monetization fuel (Story)

Others stress the difference between Google’s presentations of privacy-protecting tools at the event, most of which were “ready to ship” or “ready to demo”, with Facebook’s more aspirational approach at their own F8 event last week (“mostly vague vaporware”). Still, privacy is now at the center for both companies (Story)

Source: Artificial Intelligence on Medium

(Visited 2 times, 1 visits today)
Post a Comment