Molecularist - Part 2

What’s the fascination with humanoid robots?

Posted on 19 June 2024 by charlie

I was reading an interesting article on the fusion of robots and LLMs (see link below). One concept in the article that caught my attention was ’embodied AI’ – that current AI is ‘disembodied’ but once you ’embody’ it, the AI can learn about the world in the same way as living creatures do.

Well, not ‘living creatures’ but ‘humans’ is what article focuses on.

Dr Kendall of Wayve says the growing interest in robots reflects the rise of “embodied ai”, as progress in ai software is increasingly applied to hardware that interacts with the real world. “There’s so much more to ai than chatbots,” he says. “In a couple of decades, this is what people will think of when they think of ai: physical machines in our world.” As software for robotics improves, hardware is now becoming the limiting factor, researchers say, particularly when it comes to humanoid robots (from: Robots are suddenly getting cleverer. What’s changed?)

Hands on my body
I like ’embodied AI’ as it touches on some thoughts I’ve been having on connecting an AI with some action in the physical world. I think folks can be unfair saying that some AI is stupid, when the poor AI not only doesn’t have any connection to the physical world, but they never had the benefit of millions of years of evolution _in_ that physical world (for example, see my comment here).

So, yeah, I suppose the biologist in me groks why embodiment could do wonders for AI learning.

But then, with ‘light is over here’ lazy thinking, folks start wanting AIs to be human-smart and the navigate the world like humans. Because the world is built for humans.

Hm, the biologist in me asks ‘what’s the fascination with humanoid robots?’

BASAAP all the way down
The biologist in me sees humans as a single species among millions. And one answer to being in the real world.

Rather than say, ‘let’s make humanoid robots,’ we should be first asking ourselves (just like nature asks every day) what is the need at hand and what best addresses that need. Nature has exploded into millions of species, each evolved for their needs. The same should be for AIs embodied in robots.

Indeed, I claim that there are many tasks that humans do that would be better suited for something of a very different shape and form. Especially if that thing were clever.

For example, I wonder if a horse’s intelligence would be better for a car than a human intelligence.

It, robot
It is not that I don’t believe in humanoid robots. It’s just that I think most folks jump to humanoid robots without asking if humanoid is the right form factor.*

Decades ago** I learned the term ‘horses for courses’ – each horse has the course it is best suited for.

While I think embodied AI is a great thing to do, I just hope folks realize that embodiment can take many forms (geez, just think of all the forms manufacturing bots have).

We don’t need to ape God and make robots in our own likeness all the time. 🙄

** Frakkin’ heck, how many times in Star Wars was either C3PO or R2D2 absolutely not well suited for the environment they were in? What about Daleks (sorta)?
**Earliest reference in this blog back in 2005, tho I remember already using it in my writing in 1999.

The Narrative

Will more free time destroy the world?

Posted on 15 June 2024 by charlie

I had a brainwave [coincidentally when on my own free time holiday]. I had listened to a series of articles on long work hours in China, changes in American work hours versus others, and investigation into the national psyches behind national days. After all the talk about work vs leisure I realized that the general trend for work vs leisure time is less work and more leisure.

Industrialization, automation, AI – each wave of tech promised us we’d work less and have more time for our own pursuits. Indeed, I feel that as we make tools to take over our work load, we’ll have more attention to walk up Maslow’s Hierarchy.

Beware of what you ask for
Then, I thought back to some holidays I had, like Florence, where everything was mobbed and the sights were overrun and oversubscribed. If here in the US we gripe about travel during Memorial Day and Thanksgiving, I can’t image the gripes about travel during Chinese New Year.

Let’s just say Europe and North America start doing a three-day work week. And that the US starts making 5 weeks off normal (like lots of Europe).

Where are all the people going to go? Dubrovnik (pictured above) is always my example of over-tourism.

What if we have over-tourism all the f-ing time?

Yeah, free time is gonna kill the world. Let’s all just stay shackled to our desks.

😜

What do you think?

Strategy/The Narrative

Ghost’s manifesto is a call to take back the graph, take back the cloud

Posted on 13 May 2024 by charlie

Years ago I was trying to find away to keep folks connected across all their various digital streams. At the same time, the large Platforms were kicking off their meteoric rise that came to define the internet for the next 20 or so years. As you have no idea what I was doing back in 2007, you know who succeeded.

In the last few years, tho, people have soured on the way the internet has been built and all eyes are on the Fediverse. The Fediverse is the return to the morselization of the internet and a way to take back our social graph, take back our things that are in the cloud.

Way back when I said:

Do you want to have full control over your data, your social graph, your communications, just like you do now with your mobile phone? Source: Take back the graph! Facebook, The Cloud, and a return to the basics of social networking – Molecularist

Now, Ghost, a newsletter platform whose star has risen after the debacles at Substack, has published a manifesto of sorts in support of ActivityPub, the leading protocol for building the Fediverse.

As the article says:

This has long been the dream, and it seems like the platforms betting on it in various ways — Mastodon, Threads, Bluesky, Flipboard, and others — are where all the energy is, while attempts to rebuild closed systems keep hitting the rocks. Source: Newsletter platform Ghost adopts ActivityPub to ‘bring back the open web’ – The Verge

Alas, I really haven’t dabbled in all this Fediverse stuff. Yes, I have been banging this drum (or a version of this drum) for a while. But my head (and communities) are elsewhere these days. So, while all this is exciting for me, the Fediverse is not not really part of my life.

Maybe some other time.

Tho this quote captures the expectations I and other have as Fediverse takes flight.

That’s the fun part about the fediverse — it’s a lot of old ideas about the web being open and interoperable, but there’s still a lot of new things yet to be invented on top of that foundation. At this point I’m not sure any social platform that launches without an eye towards federation stands a chance, really. Source: Newsletter platform Ghost adopts ActivityPub to ‘bring back the open web’ – The Verge

The Narrative

Is Autopilot as smart as a horse? (another BASAAP story)

Posted on 06 May 2024 by charlie

There was an article in the Verge, discussing a recent study by the NHTSA finding Tesla’s poorly designed driver-assist features have been linked to hundreds of crashes and contributed to fatal accidents.

I’m not surprised. So much of Tesla is driven by the imagination of one man, and I keep seeing biases in their design choices that are more based on beliefs than evidence.

But one comment from that man caught my attention.

In an X (can’t call them tweets, right?), his response to a comment on autonomy is “Everything else is like variations on a horse carriage.”

Not quite betting the company, but going balls to the wall for autonomy is a blindingly obvious move.

Everything else is like variations on a horse carriage.
— Elon Musk (@elonmusk) April 16, 2024

via “Tesla’s in its flop era” – The Verge

I gotta say, if I had to choose between an autonomous car designed by Musk and a horse, designed by millions of years of evolution, I’d choose the horse.

This isn’t about being as smart as a puppy, but being as smart as a horse (#BASAAH?).

Image from DALL-E. Worked really hard to get it right and finally just chose to go with this one.

The Narrative

Adidas Sambas are over 30 years of memories for me

Posted on 19 April 2024 by charlie

I got my first Sambas (the traditional black ones, like in the photo) in college, influenced by some of my friends who were on the soccer team (bless ya, JP) .

OK, I wasn’t a footballer. But I grew up in the land of Samba and soccer, and had all sorts of Adidas stuff since a child. So there was a special connection there.

Rishi Sunak may have rendered them uncool this week, but the trainer has risen to ubiquity in the past few years. So what is it about this fairly simple design that is so widely loved? (From The highs and lows of an It-shoe: how Adidas Sambas took over the world | Fashion | The Guardian)

I wore many pairs of Sambas as my main sneaker for 30 years or so (sometimes putting in different laces)*, until the last one was worn to shreds.

Was hard to not renew my Sambas in the end. But I shifted almost 10 years ago to Vans. And now they are my main sneaker.

Interesting to see all the hullabaloo from Sunak’s reveal (and apology, it seems). Tho, I must say, the black ones I had are the only Sambas to me (which Sunak was _not_ wearing). The rest are just other Adidas sneakers, derivatives of the real deal. 😁

*Laces I remember using – hockey skate laces, purple laces, and checker laces. Style, baby, style.

Photo from Guardian article

Strategy/The Narrative

Are all GenAI buttons magic fairy dust?

Posted on 16 April 2024 by charlie

I now have apps everywhere asking me to click a button to use GenAI to do the things I can do myself. So I sorta ignore them.

But last night I saw a Samsung ad on Prime* and the GenAI photo clean up button looked quite similar to the button Evernote keeps putting in front of me.

So I did a quick search to see where else I might have seen such similar buttons.

See the gallery below for a few I quickly found in less than a minute. Aside from Samsung, these are apps I use on a regular basis.

The dream of the future
What does it say about us that all these button designs converge on stars and magic fairy dust. Indeed, that they are all too alike makes me think I am late to the party and there were some shared winks or discussions that the GenAI icon would henceforth be fanciful stars, sorta like the similarities in other icons, like ‘save’ or ‘send’.

But _even_ if there was some agreement among designers, that we chose wishful dreamy stars to represent a GenAI action betrays our techno-optimistic view of what GenAI can do for us.

Clearly the design is meant to evoke magic, inviting folks to engage in a not-threatening way (‘It’s fairy dust!’). But will it be forever, like the disk icon has remained for ‘save’ even after the last of the floppy manufacturers fade away? [Imagining Colossus with a big fairy dust button🤔]

Help out here
What do you think? Also, let me know if you see other similar buttons or fairy-like depictions of GenAI.

Now I’m going to go sprinkle some pixie-dust on this blog post I’ve been working on. 😜

*regarding Prime ad – thankfully, only at the start, and not interruptively stupid – I’m looking at you YouTube, especially during TinyDesk Concerts

Images, in case you’re still wondering, from Microsoft Copilot in Outlook MacOS, WordPress, LinkedIn, ChatGPT, Samsung, and Evernote

Strategy

Is Aboard onto something, building with GenAI? (Bonus: BASAAP reference!)

Posted on 10 April 2024 by charlie

I will admit (not that folks didn’t already know this) I am not really an original kind of guy.^[1] Tho I do fool myself into thinking I am a decent remixer, pulling together multiple threads and jammin’ a new tune. Indeed, that’s sort of what I was trained to do at the outset as a scientist – read, learn, investigate various signals, blend the threads, build models, test those models, and deliver new knowledge and insights.

I sorta never have been able to stop doing that.

Just today I was wondering (in the preverbal shower, no less) whether ChatGPT could drive a robot (or car). And then I stumbled upon this article that helped bring together to some signals coalescing in my head.

Aboard is just one of a new class of AI companies, the ones that won’t try to build Yet Another Large Language Model but will instead try to build new things to do with those models and new ways to interact with them. from: Aboard’s AI-powered bookmarking and project app is a new spin on a chatbot – The Verge

Ghost in the Machine
I feel we enjoy saying ChatGPT is only an overengieneered autocomplete because we don’t want to believe there is any intelligence in something built in such a boring way (to me: brute force linking). But if you’ve spent any time with ChatGPT (I’m conversing with it almost daily as a sparring partner, pair coder, ideation tool, among other things) there is an uncanny feeling of some sort of smarts.

Is it the smarts of a savant? Is it the smarts of a puppy (BASAAP!)?^[2] Do folks understand the concept of emergent properties?

Doesn’t matter. The interactivity and responses are frakkin’ in some uncanny zone, and only because we know how the sausage is made (sorta). And I think we’re missing the point making ChatGPT pass the Turing test. Why must we insist it be human, any more we insist a dog or a dolphin have human intelligence? We certainly suck at dog and dolphin intelligence, right?^[3] ChatGPT only needs to be a smart ChatGPT, and nothing else.

But I digress. Who frakkin’ cares? It’s a formidable tool. And as a tool, what other tricks can it do?

Invisible hand
That leads me to ChatGPT as a puppet master of some sort. What kind of front ends can we build, be they a robot or a webpage (as in Aboard), that ChatGPT manipulates (no idea how) to construct a digital output or a physical action?

If you think of it, our brain is a hyperconnected society of agents working together thru a vessel that allows us to manipulate the environment. Why can’t we provide ChatGPT some input, say converting sensors into something ChatGPT can understand (‘object 3, 2, 1 ft away’), and give it a way to manipulate an output (‘turn wheel left 2 degrees’)?

I don’t know how Aboard does the to and fro with ChatGPT. But seeing already the formats of outputs and inputs ChatGPT is already comfortable with, seeing the add-ons folks have been building to provide ChatGPT with skills, I don’t see why we can’t build other types of glue to provide ChatGPT skills to affect a user interface.

Again, I mean outputs that are not the usual textual or image or what we currently have ChatGPT do. I mean interactive things, things manipulated by ChatGPT.

Between the ears
This is just me seeing different threads and rearranging them in my head, trying to make sense of it, trying to see what permutations click (kinda like PBChoc ideation).

Dunno if this ChatGPT as manipulator is a signal or just a offshoot into nothingness. But now that I’m alert to it, Geoffrey (what I call my ChatGPT) and I will have to go out there and learn more.

[1] I once had to take a psychological exam (given by a contractor called ‘Psycho’, no less), when I was being considered for my first job at Nokia. That exam is what made me realize I was great at jammin’, just not originality. And, as a scientist and scholar, I am cool with that. Haha. Oh, and I had to take another psych exam for Business Finland too. What is it with these exams, Finland?

[2] Been thinking of digital smarts for a long time – here’s a 2016 post of mine referring to chatbot smarts, and referring back to puppy smarts from a 2007 convo on systems being BASAAP – ‘Be as smart as a puppy’. I’ve been jammin’ for a long time.

[3] Frak, a lot of footnotes today. In any case, you gotta read this book, or at least the parts on ‘alien’ intelligence. Brilliant.

Image from the Verge article

The Narrative

The connection between the T-Mobile Sidekick and Nokia Cloud

Posted on 09 April 2024 by charlie

This article got me all nostalgic for the mid-2000s. Back then, I was at Nokia, deep into the Nokia smartphone world, fusing mobile and the internet. And by 2005, looking for a new gig. I was just off of Nokia Lifeblog, having spent a few years immersed in the emerging Web 2.0 and Social Media world (we we only starting to call them that).

Before the iPhone, before Android, before webOS, a revolutionary soap bar of a phone made it incredibly easy to get shit done. The Danger Hiptop, better known as the T-Mobile Sidekick, made the internet portable and affordable like no phone before.
Source: The T-Mobile Sidekick’s Jump button made mobile multitasking easy – The Verge

Nokia Cloud v1.0
At the time, I was ruminating on an idea, partly inspired by the Danger, of not just the promise of the Web in your pocket, but the idea that the phone didn’t matter, of all your connections, streams, and stuff living in the cloud. The classic story was if your Danger fell into the drink, you’d get a new one and whoosh its previous soul would fill up the new phone – all your stuff was safe in the cloud. [Yeah, that was a big deal back then.]

I went around Nokia trying to pitch the idea of Nokia providing a place in the cloud for all your Nokia phone data and activity. We called the project Nokia Cloud, and, when launched, the service started as Ovi (‘door’ in Finnish, because I kept describing it as a door to your digital life).

I will admit I left the project just after launch, for various personal and professional reasons. While the project was an immensely intense learning experience for me, I was disappointed I wasn’t able to deliver a fuller vision.

Android
As is mentioned in the Sidekick article, many of the Danger folks went on to found Android. And my closest brush with Android was during a visit to Google a few months after Android were secretly acquired. Secretly, as Android didn’t exactly make clear what they were working on, nor did Google make a fanfare of the acquisition.

During my visit to Google, I met Andy Rubin, the top guy from Android. We had an interesting chat and then I said it was obvious they were working on a phone with Android. As far as I recall, nothing was really in the news what Android was (yes!) but I knew their pedigree and that always informed me how folks behaved.

The funny thing is that Andy got all defensive saying I was wrong. 🤷🏽 Could be he was thinking I was thinking a physical phone. Where perhaps he was thinking what I was also thinking was Danger 2.0, just the software. Keep in mind, I was also part of the launch and early days of Series 60, which was Nokia’s smartphone platform OS, sold to licensees, just like Android eventually was. So yeah, I wasn’t a phone-head but a software-head.

Alas
Android went on to own the smartphone world, only held off by Apple’s iOS. And Andy left Google with a platinum (way more than gold, y’see) parachute. Yet another one of my brushes with millionaires during that time. 🙄

And me? Well, no one is fondly remembering Ovi these days. Nokia Mobile Phones is no more. And I’m still writing this blog in anonymity. Haha.

Image from the Verge

Strategy

What’s at the heart of the iRythm Zeo XT monitor?

Posted on 05 April 2024 by charlie

I came into possession of a Zeo XT long-term heart monitor (see image right). The device was sent to someone I know who ended up not using it. So I took it and, being curious, decided to crack it open and see what I could see.

The whole device was fascinating to explore and, without the benefit of knowing the designers, contemplate the design decisions.

Cracking it open
The is the cover. Note the blue dot. That’s a button I’ll show later on. [FYI, images follow explanatory text]

The cover opened up easily enough. Later, I realized it could be when the device is sent back, the technicians need to easily open it to access a micro USB port I’ll show shortly.

The bottom that the board sits on shows where the contacts for the leads come in, and you can see that there are divots for where the contact clips connect. There are also contact clips for the batteries. Smol. Smol.

That rectangle is the recess to accommodate the memory chip.

PCB
Lots of test points. View of both sides of the board. Closeups of some areas follow.

iRhythm Tech N100A6001.

Little clippy things for leads (in right image) and for batteries (in left image).

That big chip is the ISSI IS34ML04G08. NAND memory.

Closeups
Right next to the clips for the leads (next to two large resistors) is a chip with TI 86 I333. This chips seems to be an INA to convert voltages into data.

Next to the memory chip, this chip says 263 8453 JSH. I wonder if it’s a NXP MMA8543 accelerometer. Not sure why there’s an accelerometer on this. If that’s what that chip is. I was just Googling things, not chasing it too deep.

The MCU is a ATSAM3U4C

The architecture is designed to sustain high-speed data transfers. The multi-layer bus matrix, multiple SRAM banks, PDC, and DMA support parallel tasks and maximize data throughput.

Seems like data downloading is thru USB (which MCU supports). I think that’s why it’s so easy to disassemble.

And that’s a big blue button. The cover, shown earlier, flexes to click the button. This is the logging button that the user presses (when they feel an event?). Makes sense that it’s big and sturdy.