Google and the Institutionalization of Me Too, Me Too
April 8, 2021
Never one to let a trend pass it by un-mimicked, Google has created a new YouTube feature. Ars Technica reports, “YouTube’s TikTok Clone, ‘YouTube Shorts,’ Is Live in the US.” The feature actually launched in India last September and has done well there—possibly because TikTok has been banned in that country since June. The feature but has now made its way to our shores. Writer Ron Amadeo tells us:
“The YouTube Shorts section shows up on the mobile apps section of the YouTube home screen and for now has a ‘beta’ label. It works exactly like TikTok, launching a full-screen vertical video interface, and users can swipe vertically between videos. As you’d expect, you can like, dislike, comment on, and share a short. You can also tap on a user name from the Shorts interface to see all the shorts from that user. The YouTube twist is that shorts are also regular YouTube videos and show up on traditional channel pages and in subscription feeds, where they are indistinguishable from normal videos. They have the normal YouTube interface instead of the swipey TikTok interface. This appears to be the only way to view these videos on desktop. A big part of TikTok is the video editor, which allows users to make videos with tons of effects, music, filters, and variable playback speeds that contribute to the signature TikTok video style. The YouTube Shorts editor seems nearly featureless in comparison, offering only speed options and some music.”
Absent those signature features, it seems unlikely Short will successfully rival TikTok. Perhaps it will last about as long as Stadia, Orkut, or Web Accelerator. At least no one can say Google shies away from trying things that may not work out.
Cynthia Murrell, April 8, 2021
Neuroscience To the Rescue if Developers Allow
February 5, 2021
Machine learning has come a long way, but there are still many factors that will confuse an algorithm. Unfortunately, these adversarial examples can be exploited by hackers. The Next Web offers hope for a defense against some of these assaults in, “Here’s How Neuroscience Can Protect AI from Cyber attacks.” As is often the case, the key is to copy Mother Nature. Reporter Ben Dickson writes:
“Creating AI systems that are resilient against adversarial attacks has become an active area of research and a hot topic of discussion at AI conferences. In computer vision, one interesting method to protect deep learning systems against adversarial attacks is to apply findings in neuroscience to close the gap between neural networks and the mammalian vision system. Using this approach, researchers at MIT and MIT-IBM Watson AI Lab have found that directly mapping the features of the mammalian visual cortex onto deep neural networks creates AI systems that are more predictable in their behavior and more robust to adversarial perturbations. In a paper published on the bioRxiv preprint server, the researchers introduce VOneNet, an architecture that combines current deep learning techniques with neuroscience-inspired neural networks. The work, done with help from scientists at the University of Munich, Ludwig Maximilian University, and the University of Augsburg, was accepted at the NeurIPS 2020, one of the prominent annual AI conferences, which will be held virtually this year.”
The article goes on to describe the convolutional neural networks (CNNs) now used in computer vision applications and how they can be fooled. The VOneNet architecture works by swapping out the first few CNN layers for a neural network model based on primates’ primary visual cortex. Researchers found this move proved a strong defense against adversarial attacks. See the piece for the illustrated technical details.
The researchers lament the tendency of AI scientists toward pursuing larger and larger neural networks without slowing down to consider the latest findings of brain mechanisms. Who can be bothered with effectiveness when there is money to be made by hyping scale? We suspect SolarWinds and FireEye, to name a couple, may be ready to think about different approaches to cyber security. Maybe the neuro thing will remediate some skinned knees at these firms? The research team is determined to forge ahead and find more ways to beneficially incorporate biology into deep neural networks. Will AI developers take heed?
Cynthia Murrell, February 5, 2021
Yikes! Fund People, Not Projects
January 18, 2021
“Fund People, Not Projects III: The Newton Hypothesis; Is Science Done by a Small Elite?” addresses innovation, procurement assumptions, and MBA chestnuts. The write up is long, running about 6,300 words. Here’s my summary of the argument in the research paper:
You bet your bippy, pilgrim.
Here’s the academic version of my summary:
The Newton hypothesis seems true, as far as citations are concerned: science is advanced by a small elite. This is not just “Einstein-level” breakthroughs, the small elite may not be 0.01% but 1-5% of the total number of practicing scientists. Even 10% would still cohere with the idea of scientific elitism. Citations at least on a first pass do seem to correlate with “good science” both casually (Highly cited classic papers) and by assessment of peers (Nobel prize panels; Nobel-winning papers are highly cited, and cite highly cited research).
The write up also explains why some technology organizations decline; for example, the Google. The reason is that really good people leave for greener pastures either mentally or physically. The result? Gmail goes down, Intel can’t make chips, and IBM can’t get Watson to deliver that mythical billion dollar business. Common sense, yes. Will significant change take place in staff management, procurement, or MBA thinking about innovation?
Nope.
Stephen E Arnold, January 18, 2021
A Beefed Up Elasticsearch Presages an Interesting Future
December 31, 2020
The write up “Elasticsearch New Features: 2020 Year in Review” makes several “enterprise search” issues clear:
- Key word retrieval is not enough
- Additions to basic search signals that Elasticsearch is following the Entopia, FAST Search & Transfer, and other proprietary systems down the path of exponential complexity
- Specialists in the time series and geospatial sector have cause to rejoice and be worried.
The article provides a summary of the feature landscape for Elasticsearch. It is worth pointing out that many commercial vendors rely on Elasticsearch or its cousin Lucene for information retrieval functions.
The article illustrates why. A single firm lacks the resources to build, enhance, and support what now is a retrieval and analysis platform. What’s interesting is how few vendors report their open source roots. Most prefer to concentrate on their proprietary add ons. These are the differentiators, but I must admit that most of these commercial vendors appear to me like an iguanas in a Caribbean iguana farm pen. I can no longer tell them apart. When I encounter a “new” enterprise or specialized search system positioned as a problem solver for the enterprise, I see iguanas. I suppose each iguana has a quite distinct personality, but I am not smart enough to perceive the difference.
Net net: Enterprise search is a utility. As an information service accretes features and functions, the basics become less important. At some point, the enterprise search systems, whether free or proprietary, bangs straight into the accounting department’s Zoom meeting.
The results are not pretty. Complexity, triage costs, customization costs, and special add ons set the stage for more Delphes, Fulcrums, SMARTs and STAIRS. Will vendors of enterprise search figure out how to get off this pathway to a Dante-like digital netherworld?
My prediction for 2021? Nah.
Stephen E Arnold, December 31, 2020
Innovation Shift: Not Just in China and Singapore
November 23, 2020
How much innovation will $700 million buy? I think I will find out in the next six to 12 months. “COVID Hit Startups Badly – but Something Surprising Is Happening” contains an interesting factoid, if the datum is accurate. Here’s the passage I noted:
Research from MAGNiTT, a startup data platform, revealed that $704m was invested across 564 different startups across the region in 2019. “To put it into perspective, 2009 saw $15m of funding in five venture deals,” the company noted.
The article emphasizes Covid. My thought is that other factors are contributing to this old-school Silicon Valley type surge. Regardless of the reason, some of these “deals” were ones which just a few years ago would have involved lots of Philz Coffee conversations. Not now it seems.
Stephen E Arnold, November 23, 2020
Google Reveals Its Aspiration: Everything
October 30, 2020
An online publication called Gadgets360 published “Google Renames the Chromebook Search Button to the Everything Button.” The lowly capitalization lock key has been identified as expendable. By repurposing a way to create CAPS, Google has performed two vital services:
- Easier access to search
- A way to reveal its aspiration: To be “everything” to a human user.
The article states:
Google is renaming a button on Chrome OS PC keyboards to ‘Everything Button. … Google said that the new name for the Launcher button was chosen to reflect user feedback; the search giant hoped that the inclusion of the new name for the button will help highlight that Chromebook laptops have a dedicated button on their keyboards. Clicking on the Everything Button will open up a search bar through which you can search for things on Google, as well as for apps and files on the Chrome OS machine.
Interesting. What about confusion with the freeware application called Everything. David Carpenter at Voidtools.com has offered his useful information retrieval software for several years. Google is indeed innovative and proving that it is “everything” a me-too outfit would want to be.
Stephen E Arnold, October 30, 2020
Be an Entrepreneur: The Venture Outfits Need You
October 2, 2020
Institutional Investor ran an interesting story. No, that is not an oxymoron. Really. “The Pervasive, Head-Scratching, Risk-Exploding Problem With Venture Capital.”
I noted this passage:
Two-thirds of venture deals fail, researchers have found. With such a high mortality rate, a VC fund’s actual ending portfolio size is merely one-third of its invested companies’. So to arrive at an exposure with 20 to 70 companies, a fund needs a starting portfolio of 60 to 210 startup investments. Very few funds meet this size.
Without wrestling with the assumptions or the math, I thought this statement was fascinating:
Significant portions of the average come from very few outlier deals.
Now the assertion:
The golden rule for investors into the venture asset class must therefore be: Build a portfolio of 500 startups, with 100 companies being the absolute minimum.
Okay, how many venture firms do you know that have a portfolio of 500 start ups?
Then the question, “Why not buy a fund of VC funds?” Answer:
Based on industry studies, funds of funds frequently lack diversification across gender and race.
Nervous? Not to worry. Here’s why:
Is venture capital a risky asset class? No. Most VC funds choose to act in a risky manner by not diversifying, but that does not make the asset class risky. To de-risk venture capital, CIOs simply need to acknowledge that VC math is different from public markets math. The importance of low-probability, excess-return-generating investments means that proper diversification requires a portfolio of at least 500 startups.
But most VC firms don’t have 500 or more start ups in their portfolio? That’s what the write up said.
Does this seem to be reassuring?
Stephen E Arnold, October 2, 2020
Google Decides It Is Time To Play Cards
September 8, 2020
Innovation is part of Google’s mantra. Alphabet Inc. never stops developing ideas, especially when it comes to improving its trademark product: search. Mobile search and having seamless access between mobile and desktop devices is a key selling feature. Google decided to improve its activity cards feature says Engadget in the article, “Google Promises Better Search Results For Recipes, Jobs, And Shopping.”
The activity cards feature allows users to continue searches they started on mobile devices. The feature works like this:
“Let’s say you’re looking for iPad accessories. The shopping card will display products that you’ve been researching, and even some that you haven’t explicitly searched for. If they were featured in a review or a guide, Google might surface them in the card. That could help you to compare all of your options and reach a decision.
The jobs card could make it easier for you to keep on top of new openings in your field. It’ll display relevant job listings that have popped up since you last searched, so you don’t necessarily have to trawl through the same ones over and over.”
The recipe cards work similar by keeping content on searched for recipes updated. The activity cards act like personalized RSS feeds centered on specific topics: jobs, search, shopping, and recipes. They offer a unique and customizable browsing and search option.
However, their subject reach is limited. Dozens of other apps provide the same service, but they are not limited to four topics. The only special thing about Google’s activity cards is the Google name.
How about customizing activity cards so Google users can get the most out of this feature.
Whitney Grace, September 8, 2020
Qualcomm: Tools? Nobody Had Tools.
September 7, 2020
I read “Qualcomm’s Founder On Why the US Doesn’t Have Its Own Huawei.” Interesting viewpoints appear in the article.
Here’s the passage that caught my attention:
Qualcomm, by selling companies a comprehensive chipset that could power a cellphone, actually made it easier for new Chinese competitors to hit the market, because they had the tools to create a product instantly. “Unfortunately,” he says, “nobody in the US has really run with it” and done the same thing.
And then:
Another complicating factor is that governments in China and Europe have had industrial aid policies that helped their telecom firms in a way that the US has not. “Our government has not provided R&D support or other support that Huawei and ZTE (another successful Chinese firm) managed to get from their own government,” Jacobs says.
Observation:
- US companies are as large or larger in terms of access to cash and humans. Those companies have the resources to solve certain technical problems, sustain the ecosystem to support certain technologies, and push technology which will be used by billions. What went wrong? Is it possible the US companies were supplying grain to the highest bidder. These bidders then developed industrialized farms. Looks like an error and finger pointing.
- Those who build tools often discover unexpected consequences. As a child I read science fiction books. I recall descriptions of humans building robots and then robots building robots. The story explained that humans were subjugated to their robot overlords who manufactured more and increasingly sophisticated robots. An unexpected consequence?
- “Nobody” and the “government” are convenient excuses.
Net net: Significant shift in the balance of technological power seems to be evident and industry spark plugs are not delivering the necessary oomph. Not a flattering look in what should be a public relations extravaganza, not evidence of systemic failure.
Stephen E Arnold, September 7, 2020
Interesting Insight Tucked in a Discussion of Corporate Labs
August 21, 2020
“The Death Of Corporate Research Labs” is an interesting essay. I found the references to Bell Labs fascinating. My team and I performed some small work for Bell Labs which then morphed into Bell Communication Research. Earlier I had done small work for the old Ma Bell, and it was a person from that company who submitted my name for an ASIS Award.
The write up points out that corporate research and development laboratories became a thing as a result of anti-trust pressures decades ago. Today corporate R&D is mostly a memory. Shifty eyed bean counters and Teflon coated MBAs know when to dump a cost center which does not contribute to the bottom line every 12 weeks.
The essay contains a brilliant observation. I circled this in red:
A surprising implication of this analysis is that the mismanagement of leading firms and their labs can sometimes be a blessing in disguise. The comparison between Fairchild and Texas Instruments is instructive. Texas Instruments was much better managed than Fairchild but also spawned far fewer spin-offs. Silicon Valley prospered as a technology hub, while the cluster of Dallas-Fort Worth semiconductor companies near Texas Instruments, albeit important, is much less economically significant.
Innovation is a result of lousy management.
I have to think about that because many of the high tech companies are not in my opinion well managed. Profitable? Yes. Well managed. Nope.
That raises the question:
If we accept my hypothesis that Silicon Valley high tech anti trust targets are not well managed, why are many of these firms starved for innovation?
Perhaps there is minimal correlation between management (good or bad) and innovation. The status quo suggests that me too thinking is the surest path to “innovation.”
Stephen E Arnold, August 21, 2020