You are here

Feed aggregator

Kubernetes: Good or evil? The ethics of data centers

O'Reilly Radar - Thu, 2018/11/01 - 10:00

Anne Currie says excessive and dirty energy use in data centers is one of the biggest ethical issues facing the tech industry.

Continue reading Kubernetes: Good or evil? The ethics of data centers.

Categories: Technology

Highlights from the O'Reilly Velocity Conference in London 2018

O'Reilly Radar - Thu, 2018/11/01 - 10:00

Watch highlights from expert talks covering Kubernetes, chaos engineering, deep learning, and more.

People from across the distributed systems world came together in London for the O'Reilly Velocity Conference. Below you'll find links to highlights from the event.

Kubernetes: Good or evil? The ethics of data centers

Anne Currie says excessive and dirty energy use in data centers is one of the biggest ethical issues facing the tech industry.

Incognito mentorship

Katrina Owen says the valuable skills that experienced professionals lack are at the vital margins of their careers.

Deriving meaning in a time of chaos: The intersection between chaos engineering and observability

Crystal Hirschorn discusses how organizations can benefit from combining established tech practices with incident planning, post-mortem-driven development, chaos engineering, and observability.

A new vision for the global brain: Deep learning with people instead of machines

Omoju Miller outlines a vision where we harness human action for a better future.

Learning from the web of life

Claire Janisch looks at some of the best biomimicry opportunities inspired by nature’s software and wetware.

The misinformation age

Jane Adams examines the ways data-driven recruiting fails to achieve intended results and perpetuates discriminatory hiring practices.

The freedom of Kubernetes

Kris Nova looks at the new era of the cloud native space and the kernel that has made it all possible: Kubernetes.

What changes when we go offline first?

Martin Kleppmann shows how recent computer science research is helping develop the abstractions and APIs for the next generation of applications.

Continue reading Highlights from the O'Reilly Velocity Conference in London 2018.

Categories: Technology

Deriving meaning in a time of chaos: The intersection between chaos engineering and observability

O'Reilly Radar - Thu, 2018/11/01 - 10:00

Crystal Hirschorn discusses how organizations can benefit from combining established tech practices with incident planning, post-mortem-driven development, chaos engineering, and observability.

Continue reading Deriving meaning in a time of chaos: The intersection between chaos engineering and observability.

Categories: Technology

Incognito mentorship

O'Reilly Radar - Thu, 2018/11/01 - 10:00

Katrina Owen says the valuable skills that experienced professionals lack are at the vital margins of their careers.

Continue reading Incognito mentorship.

Categories: Technology

A new vision for the global brain: Deep learning with people instead of machines

O'Reilly Radar - Thu, 2018/11/01 - 10:00

Omoju Miller outlines a vision where we harness human action for a better future.

Continue reading A new vision for the global brain: Deep learning with people instead of machines.

Categories: Technology

Four short links: 1 November 2018

O'Reilly Radar - Thu, 2018/11/01 - 03:15

Data Science, AI Ethics, Coded for Curiosity, and Worm Parking

  1. How to Decide Which Data Science Projects to Pursue (Hilary Mason) -- data science projects are not independent from one another. With each completed project, successful or not, you create a foundation to build later projects more easily and at lower cost. Some good advice on how to build a non-sucky data strategy.
  2. AI Ethics, Impossibility Theorems, and Tradeoffs -- There is no policy choice that satisfies all ethical principles. A data scientist takes us through the options and the math that makes this statement true.
  3. Reinforcement Learning with Prediction-Based Rewards (Open AI) -- We’ve developed random network distillation (RND), a prediction-based method for encouraging reinforcement learning agents to explore their environments through curiosity, which for the first time exceeds average human performance on Montezuma’s Revenge. RND achieves state-of-the-art performance, periodically finds all 24 rooms, and solves the first level without using demonstrations or having access to the underlying state of the game.
  4. C. Elegans Can Park a Car -- it only took 12 neurons, and yet you look down any city street and...*sigh*.

Continue reading Four short links: 1 November 2018.

Categories: Technology

Four short links: 31 October 2018

O'Reilly Radar - Wed, 2018/10/31 - 04:10

Who Gets What, Kindle Notes, Advertising in Young Children's Apps, and Hidden Data

  1. Rethinking Who Gets What and Why -- Tim O'Reilly's latest talk. Read the presenter notes for the meat.
  2. Klipbook -- convert highlights and notes on your Kindle to nice HTML, Markdown, or JSON.
  3. Advertising in Young Children's Apps: A Content Analysis -- Of the 135 apps reviewed, 129 (95%) contained at least one type of advertising. These included use of commercial characters (42%); full-app teasers (46%); advertising videos interrupting play (e.g., pop-ups [35%] or to unlock play items [16%]); in-app purchases (30%); prompts to rate the app (28%) or share on social media (14%); distracting ads such as banners across the screen (17%) or hidden ads with misleading symbols such as “$” or camouflaged as gameplay items (7%). Advertising was significantly more prevalent in free apps (100% vs 88% of paid apps), but occurred at similar rates in apps labeled as “educational” versus other categories. Many things happening online that were prohibited for children's TV in the 1970s. (via BoingBoing)
  4. JPG with a ZIP (Twitter) -- the image in this tweet is also a valid ZIP archive, containing a multipart RAR archive, containing the complete works of Shakespeare.

Continue reading Four short links: 31 October 2018.

Categories: Technology

Why software architects fail and what to do about it

O'Reilly Radar - Tue, 2018/10/30 - 08:00

Stefan Tilkov looks at common software architecture pitfalls and explains how they can be avoided.

Continue reading Why software architects fail and what to do about it.

Categories: Technology

Introducing serverless to your organization

O'Reilly Radar - Tue, 2018/10/30 - 08:00

Mike Roberts explores ideas for trying serverless as well as a framework for evaluating its effectiveness within your organization.

Continue reading Introducing serverless to your organization.

Categories: Technology

Career advice for architects

O'Reilly Radar - Tue, 2018/10/30 - 08:00

Trisha Gee shares advice and lessons she learned the hard way while managing her career as a developer, lead, and technical advocate.

Continue reading Career advice for architects.

Categories: Technology

Four short links: 30 October 2018

O'Reilly Radar - Tue, 2018/10/30 - 04:40

AI Animations, Dataflow Apps, Decensoring with AI, and FPGA Programming

  1. A Mixed-Initiative Interface for Animating Static Pictures -- this looks awesome!
  2. Noria: Dynamic, Partially-Stateful Dataflow for High-Performance Web Applications -- Noria makes intelligent use of dataflow beneath the SQL interface (i.e., dataflow is not exposed as an end-user programming model) in order to maintain a set of (semi-)materialized views. Noria itself figures out the most efficient dataflows to maintain those views, and how to update the dataflow graphs in the face of schema / query set changes.
  3. DeCensoring Hentai with Deep Learning -- I already don't like where AI has taken us, and we're nowhere near SkyNet yet.
  4. Spatial -- a high-level programming language for FPGAs.

Continue reading Four short links: 30 October 2018.

Categories: Technology

Potholes in the road from monolithic hell: Microservices adoption anti-patterns

O'Reilly Radar - Mon, 2018/10/29 - 11:00

Chris Richardson describes microservices anti-patterns he’s observed while working with clients around the world.

Continue reading Potholes in the road from monolithic hell: Microservices adoption anti-patterns.

Categories: Technology

The challenges of migrating 150+ microservices to Kubernetes

O'Reilly Radar - Mon, 2018/10/29 - 11:00

Sarah Wells explains how the Financial Times migrated microservices between container stacks without affecting production users.

Continue reading The challenges of migrating 150+ microservices to Kubernetes.

Categories: Technology

Are microservices a security threat?

O'Reilly Radar - Mon, 2018/10/29 - 11:00

Liz Rice outlines the security implications of microservices, containers, and serverless.

Continue reading Are microservices a security threat?.

Categories: Technology

Highlights from the O'Reilly Software Architecture Conference in London 2018

O'Reilly Radar - Mon, 2018/10/29 - 11:00

Watch highlights from expert talks covering microservices, Kubernetes, serverless, and more.

People from across the software architecture world came together in London for the O'Reilly Software Architecture Conference. Below you'll find links to highlights from the event.

The challenges of migrating 150+ microservices to Kubernetes

Sarah Wells explains how the Financial Times migrated microservices between container stacks without affecting production users.

Are microservices a security threat?

Liz Rice outlines the security implications of microservices, containers, and serverless.

Potholes in the road from monolithic hell: Microservices adoption anti-patterns

Chris Richardson describes microservices anti-patterns he’s observed while working with clients around the world.

Why software architects fail and what to do about it

Stefan Tilkov looks at common software architecture pitfalls and explains how they can be avoided.

Introducing serverless to your organization

Mike Roberts explores ideas for trying serverless as well as a framework for evaluating its effectiveness within your organization.

Career advice for architects

Trisha Gee shares advice and lessons she learned the hard way while managing her career as a developer, lead, and technical advocate.

Continue reading Highlights from the O'Reilly Software Architecture Conference in London 2018.

Categories: Technology

Four short links: 29 October 2018

O'Reilly Radar - Mon, 2018/10/29 - 04:00

Quantum Internet, Live Coding, Ethics Checklists, and Robot Compendium

  1. Quantum Internet: A Vision for the Road Ahead (Science) -- interesting paper laying out a roadmap for development of "the quantum internet." Stages: trusted repeater networks, prepare and measure networks, entanglement distribution networks, quantum memory networks, fault-tolerant few-qubit networks, and quantum computing networks. The full paper is behind a paywall (or sci-hub).
  2. Algojammer -- neat prototype of a Bret-Victor-like system to help you develop and understand algorithms. The execution of your code should be thought of as just a physical fact about the lines of text you have written. In the same way we might consider the "number of 'e' characters" in the code, or the "average line length" of the code, the "execution" of the code is just a static fact that is entirely determined by the code.
  3. Data Science Ethics Checklists -- This is not meant to be the only ethical checklist, but instead we try to capture reasonable defaults that are general enough to be widely useful. For your own projects with particular concerns, we recommend your own checklist.yml file that is maintained by your team and passed to this tool with the -l flag.
  4. IEEE Robots Guide -- a compendium of robots that are real and here today, most of which you can buy, from hands to scuttlers to humanoid robots with Einstein heads, to something that looks like a little yellow bird.

Continue reading Four short links: 29 October 2018.

Categories: Technology

Four short links: 26 October 2018

O'Reilly Radar - Fri, 2018/10/26 - 03:55

Legit DRM Hacking, CPU Emulation, Phish Yourself, and Data Structures

  1. Feds Say Hacking DRM to Fix Your Electronics Is Legal -- The Librarian of Congress and U.S. Copyright Office just proposed new rules that will give consumers and independent repair experts wide latitude to legally hack embedded software on their devices in order to repair or maintain them. This exemption to copyright law will apply to smartphones, tractors, cars, smart home appliances, and many other devices. (via BoingBoing)
  2. Unicorn Engine -- a lightweight multi-platform, multi-architecture CPU emulator framework.
  3. Gophish -- an open source phishing toolkit designed for businesses and penetration testers. It provides the ability to quickly and easily set up and execute phishing engagements and security awareness training.
  4. The Periodic Table of Data Structures -- We show that it is possible to argue about the design space of data structures. By discovering the first principles of the design of data structures and putting them in a universal model, we study their combinations and their impact on performance. We show that it is possible to accelerate research and decision-making concerning data structure design, hardware, and workload by being able to quickly compute the performance impact of a vast number of designs; several orders of magnitude more designs than what has been published during the last six decades.

Continue reading Four short links: 26 October 2018.

Categories: Technology

Machine learning on encrypted data

O'Reilly Radar - Thu, 2018/10/25 - 04:15

The O’Reilly Data Show Podcast: Alon Kaufman on the interplay between machine learning, encryption, and security.

In this episode of the Data Show, I spoke with Alon Kaufman, CEO and co-founder of Duality Technologies, a startup building tools that will allow companies to apply analytics and machine learning to encrypted data. In a recent talk, I described the importance of data, various methods for estimating the value of data, and emerging tools for incentivizing data sharing across organizations. As I noted, the main motivation for improving data liquidity is the growing importance of machine learning. We’re all familiar with the importance of data security and privacy, but probably not as many people are aware of the emerging set of tools at the intersection of machine learning and security. Kaufman and his stellar roster of co-founders are doing some of the most interesting work in this area.

Continue reading Machine learning on encrypted data.

Categories: Technology

A foundational strategy pattern for analysis: MECE

O'Reilly Radar - Thu, 2018/10/25 - 04:00

As lists are the raw material of strategy and technology architecture, MECE list-making is one of the most useful tools you can have in your tool box.

MECE, pronounced "mee-see," is a tool created by the leading business strategy firm McKinsey. It stands for "mutually exclusive, collectively exhaustive," and dictates the relation of the content, but not the format, of your lists. Because of the vital importance of lists, this is one of the most useful tools you can have in your tool box.

The single most important thing you can do to improve your chances of making a winning technology is to become quite good at making lists.

Lists are the raw material of strategy and technology architecture. They are the building blocks, the lifeblood. They are the foundation of your strategy work. And they are everywhere. Therefore, if they are weak, your strategy will crumble. You can be a strong technologist, have a good idea, and care about it passionately. But if you aren’t practically perfect at list-making, your strategy will flounder and your efforts will fail.

That’s because everything you do as you create your technology strategy starts its life as a list, and then blossoms into something else. Your strategy is, at heart, a list of lists. Thinking of your work from this perspective is maybe the best trick to creating a sane, organized, productive context for your work. Let's talk about lists for a moment.

There are two parts to a practically perfect list: it must be conceived properly, and it must be MECE, which we will define in a moment.

In a properly conceived list, two things are crystal clear:

  • Who the audience is

  • Why they care

You can determine who your audience is by asking the following key questions:

  • Upon reading this list, can the audience make a decision they could not make before having the information in the list?

  • Upon reading the list, can the audience now go do something they could not have known to do before?

These are the two reasons to bother creating any kind of information in a strategy. In this context, there is little point, time, or patience for a document that merely helps a general audience “understand” something. Your lists must be lean. That means making them directive toward work that someone will go and do, or providing the data that allows a decision-maker to decide the best course of action. The RACI (responsible, accountable, consulted, and informed) chart is a list. It answers the question for the project team of who is assigned to what role so that everyone knows who is in charge of what, who is the decision-maker for what, and who is doing the work, and if someone sees his name on the list with an "R" by an item, he can go do that work. The stakeholder list is primarily for the project manager. It lets him decide whom to include in what meetings and whom to contact for certain questions. But if these, and all the many other lists you create as part of your technology strategy, are not MECE, your building blocks will be weak and your strategic efforts will crumble. Let's look at some examples to make this clear.

This formula is MECE:

Opportunity Cost = Return of Most Lucrative Option – Return of Chosen Option

This formula is MECE:

Profit = Revenue – Cost

Revenue – Cost = Profit is MECE. That's because together those three items make a complete thought, divide across lines that don't overlap, and nothing is left out. All of the parts of the money are accounted for within the same level of discourse. It is nonsense to leave out "Revenue" and simply state "– Cost = Profit." There are only two ways to increase profit: increase revenue or decrease costs. Recognizing the formula as MECE can help remind you to address both the cost and the revenue aspects in your strategy.

This list is MECE:

Spades, Diamonds, Hearts, Clubs

This list is MECE:

Winter, Spring, Summer, Fall

Each entry in the list is mutually exclusive of every other one. There is no overlap in their content. Winter ends on a specific day of the year, and then the next day is the start of spring. Every date on the calendar is, with certainty, part of one and only one season. There is no card in the deck that is part Spades and part Diamonds.

The elements in the list, when taken together as a collection, entirely define the category. No item is left out, leaving an incomplete definition. Thus, the list is collectively exhaustive.

This is not MECE:

North, South, West

It's not collectively exhaustive. It fails to include East, and is therefore an improperly structured list.

Consider the following list:

Revenue – Cost = Profit. Free Cash Flow.

This is not MECE because "free cash flow" is not at the same level of discourse as the other items. It is true that free cash flow is an important part of any public company's earnings statements. But that is unrelated to this equation, even though they appear to all be in the category of "stuff about money in a company." That's a weak category for a list because it's not sufficiently directed to an audience for a goal.

What about this one:

Internal Stakeholders, External Stakeholders, Development Teams

This isn't MECE because "internal" and "external" divide the world between them. Development teams are a subcategory of internal stakeholders for a technology strategy.

Elements that are subcategories of other elements must not be included. Consider this list:

North, South, Southwest, East

This is not MECE because it leaves out one of the elements, West, and so is not collectively exhaustive. It also includes Southwest, which is not topologically on the same plane as the other elements. It dips into a lower level of distinction, as in the "free cash flow" example. Southwest is contained within the higher level of abstraction of South. So, the elements on this list are not mutually exclusive.

These examples are straightforward (obvious) in order to illustrate the point. But they share an attribute that precious few lists in the world have: they are enums by definition. It is clear what goes on the list and what doesn’t. Most things in life are not this simple.

Consider the following list of departments or job roles in a dev shop:

  • Software Developers

  • Architects

  • Analysts

It's not exhaustive: we left out testers, and other roles depending on your organization, such as release engineers, database administrators, project managers, and so on. To test if our list is MECE, we must ensure we have pushed ourselves to think of all the relevant components that make up that category.

Remember the first rule: know your audience. Your longer, more detailed lists should be kept for your private analysis to help you reach your conclusion, or reserved for lists of things to be done in the project, such as a work breakdown structure. But you don't want long lists when working with executives because they have executive ADD. Even though you'll worry that you're leaving out crucial things, just give them the summary, but make it MECE. Then you can reveal only the headline: the impactful conclusion that makes a difference to your audience.

The Rule of Three

A good rule of thumb is to find the level of abstraction that keeps your lists in categories of three or five items. For whatever reason, people seem to more naturally understand and remember lists of three, or at the least, odd-numbered lists. Consider two movie titles: The Good, The Bad, and The Ugly is more memorable than The Cook, The Thief, His Wife, and Her Lover. Push yourself to make your lists with three to five items. Prefer lists of three or five over lists of four. You'll find this little trick helps keep your thinking quick and nimble, and it will shorten your turnaround time because your work will be closer to what you’ll need to present to executives and stakeholders.

Consider this list of age groups:

  • 0–5
  • 6–10
  • 11–15
  • 16–25
  • 26–35
  • 36–45
  • 46–54
  • 55–65
  • 66–75
  • 76 and above

This list is technically MECE. None of the categories overlap, and the sum of the subcategories equals the whole category. It might be OK for a data scientist doing customer segmentation. But probably not even then. It’s too fine-grained and low-level, so it’s not very good for strategy work. You need to keep your visor higher; look more broadly to horizons to distill the few things that really make an impact and drive change. It’s more analysis and art than science. So, even though the list is technically correct, you will lose your audience with details like this, and you can find ways to cluster and consolidate them better, along the contours of a real difference or divergence, depending on your own organization’s products, services, and markets.

Let's look at a quick example of how to apply this idea of MECE lists.

Applying MECE Lists

Imagine you've been enlisted to create a recommendation to the CTO for a new database system to replace your legacy system. If you merely state the single database system you want to buy, any responsible executive will reject your recommendation as heavily biased, poorly considered, and potentially reckless.

So, we want to first consider our audience, with empathy, and always ask: who is this for, and what do they need to know either to make a decision or to do the thing in question?

Your deciding audience wants to know they have been given a clear, thorough, thoughtful, unbiased proposal and that they are not being manipulated. In our empathy, we realize everyone has a boss, and that no one in a company of any size just makes a decision in a vacuum. It's not the CTO's money. So your CTO must in turn answer to his bosses for the system he selects, and is accountable for its success. Your recommendation will be successful if you give your deciding audience a list of MECE lists.

But the list of database system choices is potentially in the thousands. It is impossible to include all of them, and impractical and unhelpful to include even 20 of them. Being ridiculous is not what is meant by "collectively exhaustive." So, first we'll create a list of criteria to help us make our final list MECE. Include three or five factors on which you will base your selection and write those down, as they become part of your recommendation, too. You’re showing your audience how you came to your conclusion, just like showing the long math in school: you’re not just giving the answer, but providing the steps by which you arrived at it. This helps the audience follow your story and agree with your conclusion.

Then we'll perform a survey of the landscape, including systems that meet the criteria. Include open source alternatives as well as commercial vendors. We might have a few of each. If we recommended only the one we already wanted, we would miss the chance to perform the analysis, squander an opportunity for learning that might change or augment our view, and lose confidence in our choice and ability to execute. Including only our one recommendation would certainly and immediately invite considerable skepticism and questioning about the alternatives and how we considered them.

So make a MECE list of options. The list is exhaustive according to your chosen criteria. Say you have 8 or 10 options in your list of “all the database systems considered.” Say so in your recommendation. It shows you’ve done your homework and suggests less bias and a more data-driven, analytical approach. Then say you narrowed it down to five options to present. That list includes two you reject and state why. You have a list of three options remaining.

For each element on your list of remaining recommended vendors, create another list of lists: "advantages, disadvantages" (that's a MECE list itself). The elements in each list should be something about the technology, particularly: 1) the functional requirements such as key features that distinguish it from the competition, and 2) nonfunctional requirements such as performance, availability, security, and maintainability (that's a MECE list, too). Consider these systems also from the business perspective: ability to train the staff, popularity/access to future staff, ease of use, and so forth.

Then from the list of acceptible candidates, present them all, ranked as good, better, and best. (The good, better, best list is MECE too, because you wouldn't improve its MECE-ness by adding a "horrible" option: the category or name of this list is the acceptable options, which presumably does not include "horrible," and therefore unusable ones.")

The good option might be the one that is acceptable to you, and is low cost but not optimal. The best one might be the most desirable but highest cost, and so on.

Organizing your list this way makes an executive feel more confident that you have an understanding of the entire landscape, aren’t too biased, and show your reasoning. That makes your recommendation stronger.

The Celesital Emporium...

In 1668, English philosopher John Wilkins published a proposal for adopting a universal language as well as a universal system of measurements. In his estimation, this was an entirely rational classification system.

In 1952, Argentine poet Jorge Louis Borges published an essay titled "The Analytical Language of John Wilkins." As a critique of Wilkins's work, Borges offered the following list, in his story "The Celestial Emporium of Benevolent Knowledge," purported to have been created by a 14th-century Chinese emperor as his taxonomy for classifying the members of the animal kingdom:

  1. Those that belong to the emperor
  2. Embalmed ones
  3. Those that are trained
  4. Suckling pigs
  5. Mermaids (or sirens)
  6. Fabulous ones
  7. Stray dogs
  8. Those that are included in this classification
  9. Those that tremble as if they were mad
  10. Innumerable ones
  11. Those drawn with a very fine camel hair brush
  12. Et cetera
  13. Those that have just broken the flower vase
  14. Those that, at a distance, resemble flies

The list is hilarious because it is so obviously an example of an incomplete set of sets. There's a lot left out here. Many of the categories also overlap (can a creature not be at once "fabulous" and belong to the emperor and have just broken the flower vase?). Do not all animals, at a sufficient distance, resemble flies? What belongs in "Et cetera"? Who could possibly make meaningful use of this?

Borges's point was that there is not a single, unifying, rational way to classify All The Things, that there are cultural differences that affect our views, and that ultimately such taxonomies can be shown to be arbitrary. So that's understood. The point here is that the division of animals in the "Celestial Emporium of Benevolent Knowledge" is perhaps the least-MECE list in the history of earth. Yet, how many of our project and architecture lists, on further inspection, perhaps resemble it?

Getting good at quickly checking if you are thinking in lists and then making sure they're MECE has the pleasant side effect of helping build your powers of analysis. Think of MECE as a lens. Every time you make a list, immediately test if it is MECE. Use it as a heuristic device with your team: inspect your list with the team as you’re meeting, be sure to ask if the current list you’re working on is MECE, and then refine it. Your team may groan at first, but they will gradually start to see the value, and then they will not be able to imagine how they ever lived without it.

Make your work lists of lists, and make those lists MECE. Your recommendations have a better chance of getting accepted, supported, and executed upon. And you will create more power for your organization and your team.

Continue reading A foundational strategy pattern for analysis: MECE.

Categories: Technology

Four short links: 25 October 2018

O'Reilly Radar - Thu, 2018/10/25 - 03:55

Winners Take All, Fairness, Simultaneous Translation, Secure GPUs

  1. Winners Take All: The Elite Charade of Changing the World (YouTube) -- talk at Google (!) by Anand Giridharadas, author of a book of the same name. What a great talk.
  2. Fairness and Abstraction in Sociological Systems -- Bedrock concepts in computer science such as abstraction and modular design are used to define notions of fairness and discrimination, to produce fairness-aware learning algorithms, and to intervene at different stages of a decision-making pipeline to produce "fair" outcomes. In this paper, however, we contend that these concepts render technical interventions ineffective, inaccurate, and sometimes dangerously misguided when they enter the societal context that surrounds decision-making systems. We outline this mismatch with five "traps" that fair-ML work can fall into even as it attempts to be more context-aware in comparison with traditional data science. Noted researcher (and Friend Of O'Reilly) danah boyd is a co-author.
  3. Baidu Simultaneous Translation -- audio to text, Chinese to English. Bring on the Babelfish!
  4. Graviton: Trusted Execution Environments on GPUs -- Graviton enables applications to offload security- and performance-sensitive kernels and data to a GPU, and execute kernels in isolation from other code running on the GPU and all software on the host, including the device driver, the operating system, and the hypervisor.

Continue reading Four short links: 25 October 2018.

Categories: Technology

Pages

Subscribe to LuftHans aggregator