Tuesday, December 3, 2019

AWS launches discounted spot capacity for its Fargate container platform

AWS today quietly brought spot capacity to Fargate, its serverless compute engine for containers that supports both the company’s Elastic Container Service and, now, its Elastic Kubernetes service.

Like spot instances for the EC2 compute platform, Fargate Spot pricing is significantly cheaper, both for storage and compute, than regular Fargate pricing. In return, though, you have to be able to accept the fact that your instance may get terminated when AWS needs additional capacity. While that means Fargate Spot may not be perfect for every workload, there are plenty of applications that can easily handle an interruption.

“Fargate now has on-demand, savings plan, spot,” AWS VP of Compute Services Deepak Singh told me. “If you think about Fargate as a compute layer for, as we call it, serverless compute for containers, you now have the pricing worked out and you now have both orchestrators on top of it.”

He also noted that containers already drive a significant percentage of spot usage on AWS in general, so adding this functionality to Fargate makes a lot of sense (and may save users a few dollars here and there). Pricing, of course, is the major draw here, and an hour of CPU time on Fargate Spot will only cost $0.01245364 (yes, AWS is pretty precise there) compared to $0.04048 for the on-demand price,

With this, AWS is also launching another important new feature: capacity providers. The idea here is to automate capacity provisioning for Fargate and EC2, both of which now offer on-demand and spot instances, after all. You simply write a config file that, for example, says you want to run 70% of your capacity on EC2 and the rest on spot instances. The scheduler will then keep that capacity on spot as instances come and go, and if there are no spot instances available, it will move it to on-demand instances and back to spot once instances are available again.

In the future, you will also be able to mix and match EC2 and Fargate. “You can say, I want some of my services running on EC2 on demand, some running on Fargate on demand, and the rest running on Fargate Spot,” Singh explained. “And the scheduler manages it for you. You squint hard, capacity is capacity. We can attach other capacity providers.” Outpost, AWS’ fully managed service for running AWS services in your data center, could be a capacity provider, for example.

These new features and prices will be officially announced in Thursday’s re:Invent keynote, but the documentation and pricing is already live today.

Progressive VCs and private equity are using tech and analytics to revolutionize investing

Private equity and venture capital investors are copying our counterparts in the hedge fund world: we’re trying to automate more of our job.

When I was single, I registered for (a lot of) dating websites. When I met my now-wife, I realized that any technology that can find me a spouse is a killer app. That’s why 40 million Americans use online dating sites. But, most of use raise capital and source deals the same way people looked for dates 20 years ago: networking at conferences (or bars).

Most of us want one spouse and we’re done, but in business, you want a lot of partners. I’d argue that the same type of technologies that have revolutionized dating can revolutionize our industry.

In liquid markets, most of the calories expended on technology and analytics are focused on trade selection, or “origination.” However, in private markets, there is more room to optimize across all 11 steps of the investing process. Below, I’ll walk through how progressive investors are using technology and analytics throughout all of their operations. To learn more about this space, I suggest joining an online community I co-founded, PEVCTech.

1) Managing the firm 

Before you can actually invest, you have to manage your fund. This is harder than it sounds. In the private equity universe, most partners have primary training as deal-makers, not as managers. When I talk with junior personnel at private equity firms, the quality of firm management is a frequent complaint.

I’ve used Asana extensively to manage activities firm-wide. I also use several living Google docs to maintain the minutes and the group agendas for my fixed weekly meetings. I use another live Google doc to maintain my database of companies I’m marketing to other VCs. That Google document provides cut and pasteable text I can share with other investors, based on their stage, focus and appetite.

Other investors use Trello, Basecamp, and Monday for making sure that everyone at the firm knows each others’ long-term OKRs and short-term projects. Point Nine Capital uses 15Five for continuous employee feedback.

One aspect of management which merits attention is your own cybersecurity, which should not be left until a crisis to address. Small investment firms often have interns and entrepreneurs in residence passing through, each of which is a security risk. (See A comprehensive guide to security for startups by Bessemer Ventures.)

2) Marketing

Kyle Dunn, CEO of Meyler Capital, says “investors should focus on building a large audience within a CRM system (having the ability to categorize your different constituents); communicate consistently to that audience; and implement an automation platform that can leverage lead score to profile interest. It sounds simple; however, very few asset managers actually do it.” I agree.

Many tools designed for B2B marketing in general are also relevant to investors. I know of funds using Constant Contact, Goodbits, Pardot and Publicate to create light newsletters for internal and external consumption. A major angel group uses Influitive, an advocate management tool, to track, activate and motivate their members. Other VCs use Contently* or Social Native* to create relevant content. Meyler Capital is taking the analytical rigor of modern internet marketing and applying it to fund marketing.

Point Nine Capital’s website is now powered by Contentful — it uses Unbounce for landing pages and Typeform for surveys and other data collection. “We’re using … TinyLetter for our “Content Newsletter” … and Buffer to schedule social media posts. Last but not least, we still use MailChimp to publish our (in)famous newsletter.”  I also use Mailchimp for the teten.com and pevctech.com mailing lists. Point Nine Capital uses Mention for media monitoring. Teten.com is built on WordPress as my content management system.

I use Hootsuite to coordinate my social media activity, which consists of Teten.com, PEVCTech.com, Linkedin, AngelList, and (passively) Twitter and Facebook. I use Google Drive to host my conference presentations, which are all embedded at teten.com. I use Diigo, a social bookmarking tool, to keep a record of useful websites.  I have also configured IFTTT to share on Twitter anything new I post on Diigo.

Qnary is one of numerous tools which can help build out your team members’ virtual presence. A tool like Quuu identifies relevant, shareable content to keep your social media channels active.

“There are two crucial aspects of marketing that investors often overlook: automation and analytics,” wrote Sabena Quan-Hin, Marketing Manager at Flow Capital. “Automation allows you to spend less time on tedious tasks and will help boost productivity, especially within a small marketing team. At Flow Capital, we use HubSpot’s sequences and workflows functions to automate a bulk of our emails and internal tasks. This provides us more time to develop meaningful relationships with prospects and customers. We use Google Analytics, HubSpot, and LinkedIn Campaign Manager for the majority of our analytics. For our content creation, we use tools such as Canva (graphic design) and GoToStage (webinars platform) to create and share content for prospects to find.”

3) Raising capital

Tim Friedman, Founder, PE Stack, said, “If I could offer one piece of advice to today’s managers, it would be to take the time to understand the demands of the modern institutional LP. Today’s investors are allocating more to alternatives in an environment where there are record numbers of new funds; and seeking deeper relationships with managers via direct and coinvestments. The past few years have therefore seen a huge rise in the proportion of LPs using specialized tools to manage and understand their portfolios, including platforms such as Chronograph, Solovis, Allocator, Cobalt LP, eFront Insights, iLevel, Burgiss.

The proportion of LPs using technology to manage their portfolios will continue to increase, and GPs unable to provide quality data to LPs will find it increasingly hard to retain and attract LPs. We are also seeing technology evaluation as an increasingly important part of LP operational due diligence. Excel and Google simply aren’t going to cut it if you expect to build a high quality institutional investor base.”

A more efficient approach to fundraising than haphazard networking is to mine the data exhaust from the limited partner universe to identify those LPs most likely to find your fund attractive and focus all your energy on them. I previously posted a detailed presentation with sales technology tools useful for B2B sales.

I always make a point of keeping firm records updated in the major data-trackers tracking the VC industry: AngelList, CB Insights, Crunchbase, Dow Jones VentureSource, Pitchbook, Preqin, and Refinitiv Eikon. LPs, coinvestors, and press use these tools, so I work for free for these data vendors to make sure that their data about our activities is correct. This is a great example of why data businesses have substantial moats.

Boardex and Relationship Science make it easier to understand and map social networks into potential limited partners. Cobalt for General Partners helps GPs to optimize their fundraising strategy. MandateWire and FinSearches provide leads on limited partners with new mandates which might fit your fund. Evestment is a platform for capital-raisers; Evestment TopQ automates private markets performance calculation.

I am a heavy user of DocSend, a secure content sharing and tracking platform that can be used to seamlessly share recurring materials with potential LPs. It provides analytics to track shared materials across target senders and improve the content for future leads. Point Nine Capital uses Qwilr to create modern, mobile-native collateral.

Most funds open data rooms to share previous reports, performance data, pitch decks, legal docs and other fundraising material with LPs. I’ve seen funds using Ansarada, Allvue, Box, CapLinked, dfsco, Dropbox, Digify, Drooms, Google Drive, iDeals, Intralinks, Ipreo, Merrill Corporation, and SecureDocs for their Virtual Data Rooms. These same tools are used by companies raising capital.

I’ve also experimented with using services which are marketplaces between LPs and GPs: CEPRES, DiligenceVault, FundVeil, Harvest Exchange, and Palico. Some funds are using technology-enabled intermediaries to help them sell to retail LPs, e.g., Artivest and iCapital Network.

Deer Isle Group has built the D.I.G. Beacon technology system, which automatically outbound-solicits a universe of over 10,000 institutional investors, without requiring LPs to register for an online network of funds.

Crystal guides you in how to influence a particular person, based on their online presence.  X.ai is a virtual assistant which can coordinate your fundraising and other meetings.

This 16-game arcade for AIs tests their playing prowess

Figuring out just what an AI is good at is one of the hardest thing about understanding them. To help determine this, OpenAI has designed a set of games that can help researchers tell whether their machine learning agent is actually learning basic skills or, what is equally likely, has figured out how to rig the system in its favor.

It’s one of those aspects of AI research that never fails to delight: the ways an agent will bend or break the rules in its endeavors to appear good at whatever the researchers are asking it to do. Cheating may be thinking outside the box, but it isn’t always welcome, and one way to check is to change the rules a bit and see if the system breaks down.

What the agent actually learned can be determined by seeing if those “skills” can be applied when it’s put into new circumstances where only some of its knowledge is relevant.

For instance, say you want to learn if an AI has learned to play a Mario-like game where it travels right and jumps over obstacles. You could switch things around so it has to walk left; you could change the order of the obstacles; or you could change the game entirely and have monsters appear that the AI has to shoot while it travels right instead.

If the agent has really learned something about playing a game like this, it should be able to pick up the modified versions of the game much quicker than something entirely new. This is called “generalizing” — applying existing knowledge to a new set of circumstances — and humans do it constantly.

OpenAI researchers have encountered this many times in their research, and in order to test generalizable AI knowledge at a basic level, they’ve designed a sort of AI arcade where an agent has to prove its mettle in a variety of games with varying overlap of gameplay concepts.

The 16 game environments they designed are similar to games we know and love, like Pac-Man, Super Mario Bros., Asteroids, and so on. The difference is the environments have been build from the ground up towards AI play, with simplified controls, rewards, and graphics.

Each taxes an AI’s abilities in a different way. For instance in one game there may be no penalty for sitting still and observing the game environment for a few seconds, while in others it may place the agent in danger. In some the AI must explore the environment, in others it may be focused on a single big boss spaceship. But they’re all made to be unmistakably different games, not unlike (though obviously a bit different from) what you might find available for an Atari or NES console.

Here’s the full list, as seen in the gif below from top to bottom, left to right:

  • Ninja: Climb a tower while avoiding bombs or destroying them with throwing stars.
  • Coinrun: Get the coin at the right side of the level while avoiding traps and monsters.
  • Plunder: Fire cannonballs from the bottom of the screen to hit enemy ships and avoid friendlies.
  • Caveflyer: Navigate caves using Asteroids-style controls, shooting enemies and avoiding obstacles.
  • Jumper: Open-world platformer with a double-jumping rabbit and compass pointing towards the goal.
  • Miner: Dig through dirt to get diamonds and boulders that obey Atari-era gravity rules.
  • Maze: Navigate randomly generated mazes of various sizes.
  • Bigfish: Eat smaller fish than you to become the bigger fish, while avoiding a similar fate.
  • Chaser: Like Pac-Man, eat the dots and use power pellets strategically to eat enemies.
  • Starpilot: Gradius-like shmup focused on dodging and quick elimination of enemy ships.
  • Bossfight: 1 on 1 battle with a boss ship with randomly selected attacks and replenishing shields.
  • Heist: Navigate a maze with colored locks and corresponding keys.
  • Fruitbot: Ascend through levels while collecting fruit and avoiding non-fruit.
  • Dodgeball: Move around a room without touching walls, hitting others with balls and avoiding getting hit.
  • Climber: Climb a series of platforms collecting stars along the way and avoiding monsters.
  • Leaper: Frogger-type lane-crossing game with cars, logs, etc.

You can imagine that an AI might be created that excels at the grid-based ones like Heist, Maze, and Chaser, but loses the track in Jumper, Coinrun, and Bossfight. Just like a human — because there are different skills involved in each. But there are shared ones as well: understanding that the player character and moving objects may have consequences, or that certain areas of the play area are inaccessible. An AI that can generalize and adapt quickly will learn to dominate all these games in a shorter time than one that doesn’t generalize well.

The set of games and methods for observing and rating agent performance in them is called the ProcGen benchmark, since the environments and enemy placements in the games are procedurally generated. You can read more about them, or learn to build your own little AI arcade, at the project’s GitHub page.

Xiaomi launches Mi Credit in India, a digital lending platform that offers users credit between ~$70 and ~$1,400, in partnership with ZestMoney, others (Manish Singh/TechCrunch)

Manish Singh / TechCrunch:
Xiaomi launches Mi Credit in India, a digital lending platform that offers users credit between ~$70 and ~$1,400, in partnership with ZestMoney, others  —  Xiaomi, the top smartphone vendor in India, today joined a growing wave of fintech startups in the nation that are offering credit …



Vikram lander debris: The engineer who cracked it for India

A 33-year-old space science enthusiast searched NASA moon craft images days on end to spot the elusive Indian moon lander that crashed into its surface https://ift.tt/2reU2D9 https://ift.tt/eA8V8J

Setting politics aside, Sequoia raises $3.4 billion for US and China investments

Noted Silicon Valley venture capital fund Sequoia Capital has raised nearly $1 billion for later-stage U.S. investments and roughly $2.4 billion for venture and growth deals in China, according to paperwork filed with the U.S. Securities and Exchange Commission on Tuesday.

The firm, famous for its investments in U.S. companies like Google, Instagram, Dropbox, LinkedIn, Snap and WhatsApp, is also an investor in some of China’s most successful startups.

These are companies like Alibaba, China’s e-commerce answer to Amazon; Ant Financial, a multibillion-dollar financial services powerhouse; JD.com, another e-commerce powerhouse; ByteDance, the owner of America’s latest social media sensation, TikTok; and Yitu, one of the national leaders in the development of machine learning applications.

These investments have not come without their share of controversy abroad. Yitu has been linked to the technology dragnet currently in place in Xinjiang, where an estimated 1 million religious and ethnic minorities are currently interned. Meanwhile, TikTok’s popularity in the U.S. has come with accusations of censorship in its treatment of posts that were supportive of both Xinjiang’s imprisoned population and the dissidents protesting mainland China’s increasing control over Hong Kong politics.

U.S. senators have already called for an investigation into TikTok, and Yitu was blacklisted by the U.S. Department of Commerce in October for its role in human rights violations in Xinjiang.

Setting politics aside, Sequoia has brought in $1.8 billion for its Sequoia Capital China Growth Fund V and about $550 million for Sequoia Capital China Venture Fund VII, per filings with the Securities and Exchange Commission.

It’s a sign that when valuations are concerned (ByteDance alone is now worth $78 billion, according to some reports), investors can overlook the potential political pitfalls of dealing with China.

Sequoia, led by Doug Leone, Michael Moritz, Roelof Botha and others, recently sought $8 billion for a global fund, its largest-ever fundraise, holding a first close of $6 billion in June 2018. In addition, the firm operates Sequoia Capital India, with offices in Menlo Park, Bengaluru, Mumbai, New Delhi, Singapore, Tel Aviv, Beijing, Hong Kong and Shanghai.

News of the fund comes at the tail end of another strong year for venture capital fundraising in the U.S. Firms, including 41-year-old NEA, filed to raise as much as $3.6 billion for a single fund. Meanwhile, Norwest Venture Partners, DCVC and Accel all closed new vehicles exceeding $500 million.

Fronted, from former Bud, Monzo and Apple employees, wants to make life easier for renters

Fronted, a new London-based startup aiming to make life easier for renters, is breaking cover today.

The company, founded by Jamie Campbell, Simon Vans-Colina, and Anthony Mann — former employees at Bud, Monzo and Apple, respectively — will launch early next year with a fintech product to help renters finance their rental deposits.

The plan is get accepted into the FCA “sandbox” program (run by the U.K. financial services regulator) to begin lending cash that can only be used for a rental deposit.

The thinking is that by using Open Banking and other financial technology and offering a credit product designed to finance deposits directly, Fronted can lend more cheaply than existing options, such as credit cards, pay-day lenders, and overdrafts, or insurance-backed membership schemes, and at lower risk.

“Renting sucks — anyone who rents knows it,” Fronted CEO Jamie Campbell tells me. “There are so many problems to solve and we intend to tackle them all bit by bit. But first, we are going to pay people’s rent deposits for them so they can pay us back in bite-size manageable amounts. Deposits are a large upfront expense and most people either use mum and dad to sort it out or stay where they are (in the worst cases they do to pay-day lenders)”.

In a call late last week with Campbell and CTO Vans-Colina, the pair explained that renters that apply to use the Fronted service will be asked to link their bank using Open Banking, therefore sharing their recent transaction data, and provide details of the property they wish to rent. Then, once Fronted has run the required checks and agreed to provide credit, the startup sends the money directly to the estate agent to be placed in the U.K.’s Deposit Protection Scheme, meaning that the loan never touches the renter’s hands (or wallet).

“Customers will have a direct debit to pay us back over a set schedule, or they can pay it all off when they have the money to do so, [and] we don’t charge any fees,” says Campbell. There is also a planned “holiday mode” that will allow customers to temporarily reduce their monthly payments in order to help avoid falling into financial difficulty.

“Ultimately this first product is designed to be very convenient and we believe people will opt for this more manageable alternative to a normal deposit,” adds Campbell. “There are customers of ours that will be in ‘hidden households’ unable to move because of the upfront fees… Deposits can [also] sometimes take a long time to be returned from the schemes (something the government recently launched an enquiry into). Fronted wants to serve people who might otherwise be ‘double-exposed’ by deposits. We hope this first product increases social mobility by providing liquidity when people need it”.

Initially, Fronted will generate revenue through interest charged. It then plans to extend its fintech product offering with additional money-advance services “to help smooth out the bumps of renting”.

“We also intend on rolling out a ‘turn up and turn on’ service for utilities and internet,” says the Fronted CEO.

Plant-based burgers are “ultra-processed” like dog food, meat-backed ads say

The White Castle Impossible Slider is significantly heftier than the White Castle Beef Slider.

Enlarge / The White Castle Impossible Slider is significantly heftier than the White Castle Beef Slider. (credit: Eric Bangeman)

A public-relations firm backed by meat producers has unleashed a savage marketing campaign that claims plant-based meat alternatives are unhealthy, "ultra-processed imitations" similar to dog food.

The campaign rolled out in recent weeks from the industry-funded firm Center for Consumer Freedom, according to The New York Times. So far, it has included full-page ads and opinion pieces in mainstream newspapers, including The New York Times, USA Today, and The Wall Street Journal. All the marketing material raises health concerns about trendy meat alternatives, such as the Impossible Burger and Beyond Burger.

One ad posed the question "What's hiding in your plant-based meat?" Another directed readers to take the quiz "Veggie Burger or Dog Food?"

Read 9 remaining paragraphs | Comments

https://arstechnica.com

Late-stage Indian startups bulk up its boards with independent hands

Oyo, Ola, Freshworks, Flipkart and Paytm are among leading home-grown technology ventures that have brought in known industry veterans over the course of the year. https://ift.tt/360GJ88 https://ift.tt/eA8V8J

Consumers will soon be able to use UPI for cash withdrawals

Public sector lender Bank of India has introduced this facility for its own customers, in partnership with Mumbai-based payments company AGS Transact Technologies. https://ift.tt/33HyNal https://ift.tt/eA8V8J

Monday, December 2, 2019

Wide-ranging interview with AWS CEO Andy Jassy on the future of the cloud, the competition, market shifts, changing customer requirements, and more (John Furrier/SiliconANGLE)

John Furrier / SiliconANGLE:
Wide-ranging interview with AWS CEO Andy Jassy on the future of the cloud, the competition, market shifts, changing customer requirements, and more  —  In an exclusive one-on-one conversation, Amazon's cloud chief reveals how he views the future of the cloud, the competition, market shifts, customer demands and controversies



Mass media vs. social media

In the waning years of the last millennium, at my university, one of the cause célèbres of the progressive left was a concept known as “Manufacturing Consent,” the title of a book and film, by and starring Noam Chomsky. Its central thesis was that US mass media “are effective and powerful ideological institutions that carry out a system-supportive propaganda function, by reliance on market forces, internalized assumptions, and self-censorship.”

It’s fair to say that history has been pretty kind to this theory. Consider the support drummed up by mass media for the invasion of Iraq in 2003. To quote the public editor of the New York Times, “To anyone who read the paper between September 2002 and June 2003, the impression that Saddam Hussein possessed, or was acquiring, a frightening arsenal of W.M.D. seemed unmistakable. Except, of course, it appears to have been mistaken.” Consider the September 2002 dossier published by the UK government “to bolster support for war” which turned out to be full of spectacularly incorrect information, and the media’s failure to interrogate those claims.

It’s hard to overstate just how cataclysmic these errors were. If the mass media had pushed back against the false claims of weapons of mass destruction, we might have avoided the Iraq war, which killed hundreds of thousands and cost trillions of dollars. Saddam Hussein was not exactly a tough act to follow, but the US still managed to follow its falsely motivated war with a botched occupation which turned Iraq, and arguably the larger Middle East to this day, into a bloodbath.

An interesting question is: what would have happened if today’s social media had been around in 2003? Today, if a wrong assertion is promoted by the mass media, it doesn’t take long for subject-matter experts to appear on Facebook and Twitter, correcting them, and either going viral themselves or becoming the subjects of countervailing media stories.

This doesn’t necessarily mean catastrophe would have been averted. But at least a possible corrective to the collective hysteria of the mass media would have existed, unlike in 2002-3. (Yes, those were the days of Blogspot and LiveJournal, but they didn’t have anything like the reach or significance of today’s social media.)

Consider a more recent event: the 2016 American presidential election. It has become an article of faith, in certain quarters, that it was won and lost by the diabolical use of Facebook ads, especially in conjunction with the psychographic superscience of Cambridge Analytica. This is ridiculous. First, no one credible thinks CA’s purported ability to mind control Facebook users by showing them “psychographically” targeted ads was anything other than snake-oil nonsense.

Second, as Nate Silver points out, the impact of social-media ads was enormously less than the impact of mass media. Remember the months of hysteria about Hillary Clinton’s emails? Remember how it turned out to be a complete non-story? Doesn’t this remind you of Iraq’s WMDs?

“The media’s coverage of Hillary Clinton’s email scandal was probably literally 50 times more important to the outcome of the 2016 election than Trump ads on Facebook.” Perhaps, my fellow mass media, the fault lies not in our psychographic bullshit artists, but in ourselves

Social media has many downsides. You don’t have to go particularly deeply into my own back catalog to discover that I am a harsh critic of Facebook myself. But let’s not pretend that mass media, just because it’s older, is therefore perfect. It has its own catastrophic failure modes itself. In fact — whisper it — maybe we’re a lot better off, net, with social media and mass media, in that each can act as a counterbalancing corrective on the other’s flaws and failure modes.

The progressive left may have gone from “mass media is the enemy” to “Big Tech social media is the enemy,” but maybe, and I know this sounds crazy because it’s on the Internet, but hear me out here, maybe there’s room for a little nuance; maybe they both have good and bad aspects, and could possibly balance one another out. If you don’t think mass media needs a corrective, let me remind you once again of the Iraq War and But Her Emails, to name but two of many, many examples. Maybe there exists a future in which social and mass media are each a cure for what ails the other.

FBI wants Americans to delete these apps from their phones

The Federal Bureau of Investigation considers any mobile app developed in Russia to be a "potential counterintelligence threat," it said on Monday, responding to a U.S. lawmaker's query about face-editing photo app FaceApp. https://ift.tt/383PQqy

Anthropic updates Claude to add custom writing styles tailored to users, aiming to personalize its replies, following similar ChatGPT and Gemini features (Jess Weatherbed/The Verge)

Jess Weatherbed / The Verge : Anthropic updates Claude to add custom writing styles tailored to users, aiming to personalize its replies,...