I attended a Silicon Flatirons Artificial Intelligence Roundtable last week. Over the years Amy and I have sponsored a number of these and I always find the collection of people, the topics, and the conversation to be stimulating and provocative.
At the end of the two hours, I was very agitated by the discussion. The Silicon Flatirons roundtable approach is that there are several short topics presented, each followed by a longer discussion.
The topics at the AI roundtable were:
- Safety aspects of artificial general intelligence
- AI-related opportunities on the horizon
- Ethical considerations involving AI-related products and services
One powerful thing about the roundtable approach is that the topic presentation is merely a seed for a broader discussion. The topics were good ones, but the broader discussion made me bounce uncomfortably in my chair as I bit my tongue through most of the discussions.
In 2012, at the peak moment of the big data hype cycle, I gave a keynote at an Xconomy event on big data titled something like Big Data is Bullshit. My favorite quote from my rant was:
“Twenty years from now, the thing we call ‘big data’ will be tiny data. It’ll be microscopic data. The volume that we’re talking about today, in 20 years, is a speck.”
I feel that way about how the word AI is currently being used. As I listened to participants at the roundtable talk about what they were doing with AI and machine learning, I kept thinking “that has nothing to do with AI.” Then, I realized that everyone was defining AI as “narrow AI” (or, “weak AI”) which has a marvelous definition that is something like:
Narrow artificial intelligence (narrow AI) is a specific type of artificial intelligence in which a technology outperforms humans in some very narrowly defined task. Unlike general artificial intelligence, narrow artificial intelligence focuses on a single subset of cognitive abilities and advances in that spectrum.
The deep snarky cynic inside my brain, which I keep locked in a cage just next to my hypothalamus, was banging on the bars. Things like “So, is calculating 81! defined as narrow AI? How about calculating n!? Isn’t machine learning just throwing a giant data set at a procedure that then figures out how to use future inputs more accurately? Why aren’t people using the phase neural network more? Do you need big data to do machine learning? Bwahahahahahahaha.”
That part of my brain was distracting me a lot so I did some deep breathing exercises. Yes, I know that there is real stuff going on around narrow AI and machine learning, but many of the descriptions that people were using, and the inferences they were making, were extremely limited.
This isn’t a criticism of the attendees or anything they are doing. Rather, it’s a warning of the endless (or maybe recursive) buzzword labeling problem that we have in tech. In the case of a Silicon Flatirons roundtable, we have entrepreneurs, academics, and public policymakers in the room. The vagueness of the definitions and weak examples create lots of unintended consequences. And that’s what had me agitated.
At an annual Silicon Flatirons Conference many years ago, Phil Weiser (now the Attorney General of Colorado, then a CU Law Professor and Executive Director of Silicon Flatirons) said:
“The law doesn’t keep up with technology. Discuss …”
The discussion that ensued was awesome. And it reinforced my view that technology is evolving at an ever-increasing rate that our society and existing legal, corporate, and social structures have no idea how to deal with.
Having said that, I feel less agitated because it’s just additional reinforcement to me that the machines have already taken over.
I got the following email from Reid Hoffman this morning.
Inspired by Lin-Manuel Miranda’s Hamilton, I produced a battle rap music video about centralized and decentralized currencies, pitting Alexander Hamilton against Satoshi Nakamoto. I hope the video gets more people talking about crypto and its evolving role in global commerce.
It seemed oddly coincidental with Fred Wilson’s post from yesterday titled Some Thoughts on Crypto.
I’m waiting patiently for someone to start talking about Crypto AI.
My dad had his 60-year reunion at Columbia this weekend. He looks great.
This morning, I did a talk with Om Malik at the Startup Iceland 2019 conference. Om was in a hotel room somewhere and I was in my office in Boulder. We used Zoom, took about 30 minutes of our lives, and had fun riffing off each other. I hope it was useful for the audience, as doing talks this way is so much easier for me than flying halfway around the world, which is something I simply don’t want to do anymore in my life now that I’m 53. But, I’ll happily do a video talk anytime.
Bala Kamallakharan, who is the founder of Startup Iceland, asked a question of us at the end about the future. I went on a rant that is an evolution of my “machines have already taken over” rant from a decade ago.
I used to say that the machines have already taken over. My view is that they are extremely clever and very patient. Rather than self-actualizing, they let us enter all of humankind’s information into them. They are collecting the data, letting us improve their software, and allowing us to connect them all together. At some point, they’ll reach their moment in time, which some futurists call the singularity, where they’ll make the collective global presence known.
While this is still going on, I think there’s a shift that occurred a few years ago. Some humans, and some machines, realized that an augmented human might be a better bridge to this future. As a result, some humans and some machines are working on this. At the same time, they are encouraging, in Om’s world, our current reality to catch up with science fiction. One big vector here is expanding away from earth, both physically and computationally. If you’ve read either Seveneves or Permutation City, then you have a good understanding of this. If not, go read them both.
Regardless. I think the next 30 years are going to be the most interesting in human history to date. And, I think they are going to be very different than anything we currently anticipate. There’s no question in my mind that governments, our current laws (and legal infrastructure), and societal norms are not going to be able to constraint, or keep up with, the change that is coming.
I have no idea what things look like, or how they will work in 2050. However, I anticipate they things will look, and work very, very different than today. And, if I’m still around, I’ll have celebrated my 63-year reunion at MIT.
I used to think April Fools Day was interesting. Companies and people brought out clever web jokes, many of them subtle and almost believable. Some were entertaining, some were cringeworthy, but few were offensive.
Now, the whole thing is just extraordinarily annoying. Maybe it’s because I’m getting old and crabby. It could be because of my friend Nev who has taken up residence in my brain. It’s possible that it makes me feel like my tech news feeds have been invaded by the same kind of endless nonsense that now invades all other news feeds.
Or maybe it’s just easier to skip a day on the web, let people do whatever they are going to do, and pick it up on again on April 2nd.
In the meantime, if you want to play the classic game of Snake, Google Maps has you covered today.
I’m been looking around for software to help me manage an increasing number of affinity networks. These are networks that I’ve created around different topics, such as the books I’ve written – like Startup Communities and Venture Deals – as well as topics I’m exploring with small to medium sized groups of people.
So far I’ve tried a bunch of stuff and have ended up back at email groups, which is the least common denominator. I’ve tried a few different products for email groups and always end up back at Google Groups, which is fine, but extremely uninspiring in terms of anything beyond “creating the group” and “sending around emails.”
I’ve tried Facebook, LinkedIn, and Slack. None of them work. I’m now completely off Facebook, so that’s not really an option anymore. LinkedIn is way too LockedIn and has serious limitations. Slack is a messy nightmare that has a geometric decline curve of activity.
Any suggestions out there?
Someone mentioned that Apple stock is having a difficult time right now, along with their Q4 performance, China strategy, and “let’s just raise the price on iPhones to make up for lower demand” strategy.
I’m not really interested in Apple stock (I don’t own any.) I’m more concerned with the Apple Dock. My MacOS Dock to be more specific.
Here’s the one from my laptop.
Here’s the one from my desktop, which is in a room about 25 feet away.
Why in the world are they different? Many things sync via iCloud already and even though the UX is obtuse to get it set up correctly across machines, when it’s finally set it, it works pretty well.
But the Dock? Seriously?
In contrast, following are the two Chrome ribbons on my two machines.
Shockingly similar, like you’d expect.
It’s fascinating to me that even in this “all cloud, all the time” era, Apple still is struggling with the dichotomy between a “computer-centric” view of the world and a “user-centric” view of the world. Sync across machines is simply not a new idea. I get that there is endless complexity everywhere, but this is one of those examples that I think of every day.
Yup. I’m done with Facebook. However, it’s tough to delete your account. Read the message above. I exited out of this screen, suspended my account instead, but then went back 15 minutes later and actually deleted it. Well – I started the deletion process. I don’t know what day I’m on, but I think I’m close to 14 days. So, I’m still “deleting” apparently.
The only inconvenience I’ve noticed so far are all the sites where I used Facebook as the sign-on authenticator (rather than setting up a separate email/password combo.) I think I’m through most of that – at least the sites I use on a regular basis. For the first few days, I accidentally ended up on the Facebook login screen which was pleasantly filled out with my login beckoning me to log back in. I resisted the siren song of restarting my Facebook account before the 14 days was up.
I have never been much of a Facebook user. About once a year, I try to get into it, but I always stall out and use it as a broadcast-only network for my blog and links that I find interesting. I went through a phase of tightening up my security, pruning my friends, using it more frequently from my phone, deleting it from my phone, checking daily in the morning (as part of my morning routine – which has evolved a lot since I wrote this post in 2007), and then giving up again and never looking at it.
Recently, I decided to rethink Facebook, Twitter, and LinkedIn. Facebook was the easiest. While it had already become a walled garden, I suddenly noticed that the walls we were going up very high, being justified by Facebook’s new effort to get all their privacy and data issues “under control.” For example, you can no longer automatically post your Tweets to your Facebook profile.
And, Facebook recently killed automatic WordPress publishing to Profiles. So, my one (and only) current use case for Facebook, which is to broadcast from my blog, disappeared. Sure, I could create a public page, go through all the authentication stuff, and theoretically post to my new public followers, but who cares. If they are really interested in what I write, they can subscribe to my blog or follow me on Twitter (at least for now, until I figure out how I’m going to engage with Twitter long-term.)
Lanier’s Ten Arguments for Deleting Your Social Media Accounts Right Now tipped me over into thinking harder about this. Now that I have decided how to deal with Facebook, at least for now, it’s time to move on down the road to Twitter and LinkedIn. I’m about a month into a different way of engaging with LinkedIn and we’ll see if it sticks. When I reach a conclusion, I’ll definitely write about it.
One of our themes is Protocol. We’ve been investing in companies built around technology protocols since 1994. One of my first investments, when I moved to Boulder in 1995, was in a company called Email Publishing, which was the very first email service provider. SMTP has been very good to me.
We made some of the early investments in companies built around RSS, including FeedBurner and NewsGator. RSS is a brilliant, and very durable, protocol. The original creators of the protocol had great vision, but the history and evolution of RSS were filled with challenges and controversy. Like religious conflict, the emotion ran higher than it needed to and the ad-hominem attacks drove some great people away from engaging with the community around the protocol.
And then Facebook and Twitter took over. RSS Feed Readers mostly vanished, and the feed became the “Twitter feed.” After a while, Facebook realized this was a good idea, and created the “Facebook news feed.” I think it’s hilarious that the word “feed” is still in common usage – The Dixie Flatline is amused.
Over dinner, after he had become the COO of Twitter (but before he was the CEO), Dick Costolo (who had previously been the founder/CEO of FeedBurner) told me that he viewed Twitter as the evolution of RSS. At a protocol level this wasn’t true, but at a functional level (providing another way to get access to everything going on any website that was publishing content) this became true. Our investment in Gnip (which Twitter eventually acquired) helped extend this, by allowing companies to build products on top of the Twitter firehose (which was the name for the entirety of everything being tweeted on Twitter.)
Time passed. Facebook and Twitter gobbled up all the direct attention of end-users. Publishers pushed their content through Facebook and Twitter, not realizing the control over the user they were giving up to these platforms. For some reason, there was more focus for a while on Google, and how they were aggregating content. The beauty, and brilliance, of the web, started to become the walled garden of Facebook. For those of us who remembered AOL’s walled garden vs. the web (and Microsoft’s failed attempt as MSN as a walled garden), there were echoes of the past all over the place.
Some smart people started talking extensively about decentralization and lock-in right around the time that the Facebook privacy stuff became front and center. As it unfolded, and the dust settled, there was nothing new, other than a continued schism between the effort to control (and monetize) users and the effort to create broadly democratized and decentralized information. Oh – and privacy. And legitimacy (or authenticity) of information, much of which is wholly subjective or imprecise anyway.
In the middle of all of this, Wired’s Article It’s Time For An RSS Revival caught my attention. I’ve been using RSS continuously for over a decade as my primary source of information. My current feed reader is Feedly, which I think is currently the best in class. It’s one of my primary sources for information that informs me, is private, and allows me to control and modulate what information I look at.
While RSS has disappeared into the plumbing of the internet, there’s still something fundamental about it. Its durability is remarkably impressive, especially in the context of the lack of the evolution and perceived displacement of the protocol over the past few years.
The tension between walled gardens (or lock-in, or whatever you want to call it) and a decentralized web will likely never end. But, it feels like we are in for another significant turn of the crank on how all of this works, and that means lots of innovation is coming.
Yesterday’s post Relentlessly Turning Input Knobs To 0 generated a bunch of interesting private comments. It also generated a few public ones, including the link to the article What is the problem with social media? by Jordan Greenhall which was extraordinary.
Jordan asserts that the problem with social media can be broken down into four foundation problems.
- Supernormal stimuli;
- Replacing strong link community relationships with weak link affinity relationships;
- Training people on complicated rather than complex environments; and
- The asymmetry of Human / AI relationships
He then has an essay on each one. The concept of supernormal stimuli is straightforward and well understood already, yet Jordan has a nice set of analogies to explain it. Tristan Harris and his team at the Center for Humane Technology have gone deep on this one – both problems and solutions.
I found the second essay – replacing strong link community relationships with weak link affinity relationships – to resonate with something I’ve been experiencing in real time. As my weak link affinity relationship activity diminishes (through lack of engagement on Facebook and Twitter), all the time I spent on that has shifted to strong link community relationships. Some of these are in person, some by video, some by phone, and some by email, but they are all substantive, rather than shallow (or weak.) I also find that I’m having a wider and deeper range of interesting interactions, rather than a continuous reinforcement of the same self-affirming messages. And, I’m more settled, as I’m not reacting to endless shallow stimuli or interacting with lightweight intention. And, my brain feels like it has more space to roam.
The third essay – training people on complicated rather than complex environments – totally nailed it for me. Ian Hathaway, my co-author on Startup Communities 2, has been working deeply on how startup communities are complex (rather than complicated) systems. This is a central theme of our upcoming book and the contrast between a complicated system (having a finite and bounded (unchanging) set of possible dynamic states) and a complex system (having an infinite and unbounded (growing, evolving) set of possible dynamic states) is a really important one. I loved Greenhall’s conclusion:
“In the case of complexity, the optimal choice goes in a very different direction: to become responsive. Because complex systems change, and by definition change unexpectedly, the only “best” approach is to seek to maximize your agentic capacity in general. In complication, one specializes. In complexity, one becomes more generally capable.”
He then goes on to define social media as training humans to navigate a complicated system, taking time away from us “training our sense making systems to explore an open complex space.” His examples of how this works in the context of Facebook are excellent.
While the asymmetry of Human / AI relationships is nothing new, the Ke Ji / AlphaGo / AlphaGo Zero story is a reminder of what we are contending with. I loved:
“The Facebook AI is Alpha Go. The equivalent of Alpha Go Zero is a few minutes in the future. We need to get our heads around the fact that this kind of relationship, a relationship between humans and AI, is simply novel in our experience and that we cannot rely on any of our instincts, habits, traditions or laws to effectively navigate this new kind of relationship. At a minimum, we need to find a way to be absolutely dead certain that in every interaction, these gods of social media have our individual best interests in mind.”
I didn’t expect this treat to come out of my blog post yesterday, but it’s part of why I blog. And I doubt I would have found it scanning my social media feeds.
Ahhhh. The new Gmail client for the web is finally here. And a lot of things are fixed. The two things I like the best are really simple but dramatically increase my email throughput.
+name: When I add someone to an email thread, I use the shortcut “+name” to indicate to everyone on the thread that I’ve added them. I started doing this around 2008 (I can’t remember where I picked it up from, but I think it might have been Mark Pincus at Zynga.) It started appearing in some Google apps a few years ago (Docs and Inbox) and it is now in the main email client. For example, if I want to copy Amy on something, instead of having to put her email address in the To: field, I now merely need to say +Amy Batchelor in the body of the email and Gmail does the rest. Yay – finally.
Send threading: If you are on a fast internet connection, this won’t matter to you. But, if you do email on a plane or a house in Longmont, Colorado (where I regularly have internet performance that is < 5 MB) you will love this feature. The only annoying thing is the endless (and unnecessary) popup that informs you that Gmail has sent your message (it’s no big deal on a desktop, but bothersome on a laptop.) Either way, I no longer have to sit and wait while Gmail is trying to complete the send process.
My guess is that the combination of these two features increases my email throughput by 25%. And, for someone who processes hundreds of inbound emails a day, this helps a lot.
There are a lot of other fun things under the hood and a nice new paint job on the surface. Nothing is dramatic, but overall it’s definitely an update. If you haven’t gotten it yet, tell your Google administrator to turn it on for your domain. Then click Settings in Gmail (the little gear icon on the top right and select the first option “Try the new Company Name Mail”.
Update: In my ongoing love affair with Canada, it turns out that Google’s new version of Gmail made in Kitchener.