The Simplicity-Frugality Quad

In this blog post, we cover the philosophical underpinnings of HAL51 AI manifested in the form of the Simplicity-Frugality quad as seen in the figure above with the constituent nodes being: The Least Squares-Pepper’s Ghost nexus, The Tao of Daiso, Retrofuturism and The Binocular Curse.

We now expand upon these ideas in detail below.

1: The Least Squares-Pepper’s Ghost nexus

Humanity grapples with a weird pro-complexity bias. Simple elegant solutions are often overlooked in favor of bloated complicated overkills. An archetypal example of this is the Least Squares paradigm. In another life, we were repeatedly shocked while working in physical layer wireless communications (specifically, channel estimation), how often simple least squares and her cousins (The Tikhonov regularized hack and the Shrinkage family) would just work and the added complexities and latencies that came up the new kids on the block (read the MMSE family) was often not worth the effort at all. Yet, we saw this constant push-back, purely because the boffins felt that Least Squares was too simple and too auld and was championed in the 1700s/1800s by the likes of Cotes, Laplace, and Legendre.

Do you know what else burst on the scene in the same-ish time period? A hauntingly beautiful and importantly, convincing, illusion called the Pepper’s Ghost. (Giambattista della Porta in 1584 had mentioned a similar trick in his treatise Magia Naturalis). It was the quintessential *19th-century parlor trick used in Victorian stagecraft.* Legend has it that the illusion was so convincing that after one particularly memorable exhibition, a stunned Michael Faraday approached its inventor, John Henry Pepper, and timidly quipped:  “Do you know, Mr. Pepper, I really don’t understand it!” .

As time passed, it became phenomenally successful around the globe and has captivated audiences at mega-events such as the 2012 Tupac hologram concert at Coachella and the 2014 Michael Jackson hologram concert at the Billboard Music Awards. Recently however, a bizarre combination of misplaced hardcore-ism by phase-fetishizing hologram-pundit puritans and a few scammy entrepreneurs who keep peddling holo-x named Kickstarter projects (such as Holus and Holovect), it has become taboo to associate terms such as ‘Hologram’ and ‘3D’ with any device that utilizes the Pepper’s Ghost. Well, we at HAL51.AI have no place for such infantile parochialisms and childish stigmas. In our viewpoint, it is precisely the same kind of dogmatic gatekeeping that the Kernel-Machine/Max-Margin-Uncles kept neural networks relegated as some kind of an empirical gimmick. Plus, we always root for the underdogs anyway (Go Nottingham Forest! Vivit Post Funera Virtus). This brings us to the next node of the quad: The Tao Of Daiso.

2: The Tao Of Daiso (Psst.. it’s not a Japanese Dollar Store)

Most consumers tend to miss the underlying philosophy of Daiso. There is this misplaced anchoring of the Daiso phenomenon as the quintessential low-cost Tchotchke - Trinkets - Bric-a-bracs business play. Nope! That would be your garden variety Dollar store hellscape (Thanks John Oliver for enlightening us about how bad they really are!). As the world bid adieu to Hirotake Yano earlier this year, the prodigal 100 Yen Man and founder of the Daiso conglomerate, two of his legendary quips began to trend in the business philosophy worlds via his obituaries that aptly captured what the underlying nous of Daiso was all along:

1. "It would be all over if the customers became bored”

2. “Always pursue the excitement of discovery. Sell something unexpected!”

Through the Daiso movement, he showed that getting a consumer to breathe rarefied air at the peak of Maslow's Hierarchy Of Needs can be reasonably achieved at 100 Yen! In fact, the literal first line of their corporate website is: “Our mission is that every one of our stores provides a unique ‘surprise & delight’ shopping experience”.

The qualia of the catch-phrases “Surprise and Delight” and “Sell something unexpected” strongly capture a big facet of who we are and what our worldview of the consumer-tech world is. And yes! One day, we do unapologetically hope to get into the Diasos of the world.

3: The Binocular Curse

There’s an apocryphal tale we heard about a hardware entrepreneur who made a killing selling entry-level telescopes during the pandemic. When asked who his biggest competitor was, the one he feared the most, he apparently admitted: “NASA!” As it turns out, NASA writes beautiful blog posts educating citizen scientists (his bread-and-butter consumer base) on the best practices for viewing the night sky. In one such blog post they declared that

• “A good pair of binoculars, ranging in specifications from 7x35 to 10x50, will give you great views of the Moon, large open star clusters like the Pleiades (M45), and, from dark skies, larger bright galaxies like the Andromeda Galaxy (M31) and large nebulae like the Orion Nebula (M42)”.

So the hardware entrepreneur knew his clientele really well. They were citizen-science-loving retirees and tech nerds and there was a huge overlap between this star-gazing community and the bird-watching/birding community. Realizing that the pair of binoculars that they owned for birding was possibly superior to the low-end telescope they were about to purchase from the entrepreneur’s online store would sound a death knell to a significant chunk of his business. While this seems too reductionist and an over-simplified tale, it does teach us a deeper lesson. That, driven by hyper-consumerism, a huge chunk of us tech consumers often end up hoarding a whole lot of devices and gadgets with overlapping or worse, sub-set functionalities, just cloaked in different form factors. This is exactly what we felt about a lot of “AI-devices” that have recently sprung up, many of which are cloaking slices of silicon-real estate that are strictly inferior to the one already sitting in your pocket or catching dust in your cupboard in the form of a smartphone. Want an SBC (Single Board Computer) with a small-GPU that’s running a Linuxy open-sourcy OS that can potentially also run a tiny LLM or can reliably connect to one running in the cloud? Yeah. That’s your smartphone 😅

Why need an Orange Pi / A Khadas or a beefier cousin of the ESP-32 stable when you have a Pixel Phone in your possession? Much akin to the binocular add-ons such as bespoke tripod mounts and lenses that transmorph them into telescopes, we believe that a similar analogical play emerges for the Smartphone-3D-AI triad leading to our core thesis point: The institution of the AIccessory!

4: Retrofuturism (And why now?!)

Here are two factoids about us:

1. We are burners

2. We are Steampunk aficionados

Burn after burn, we’ve witnessed and enjoyed the incredible Steampunk art brought to the play that instantly captured our attention. So, yeah. The anachronism and the retro-futurist vibes have always had a natural sense of allure for us. Hence, when the dream of infusing life into the ghosts that haunt a petite Pepper’s Ghost box picked straight out of the Victorian theater from the 18th century using the SoTA LLMs of 2024 became a real possibility, we jumped onto it. While there are dime-a-dozen projects on Kickstarter that have been attempted in the past, this time, the vibes hit differently. Even though Siamese nets, LSTMs, RNNs, and CNNs all existed in the 1980s, it required NVIDIA’s GPU muscle power + Big Data to bring scale into the mix and thaw the AI winter. For the Pepper’s Ghost-AI nexus, there are not one, not two, but three NVIDIA-shaped perfect storms pushing the ship forward: The 3D computer vision revolution (See figure below), cheap and awesome OLED screens on the phone, and the commodification of LLMs to provide both knowledge and autonomy to the ghosts. We believe that those entrepreneurs on Kickstarter did nothing wrong. They were a tad bit too early. And hence Marc Andreessen has indeed been proven right again: “There are no bad ideas in Tech. Only bad timing:”

The landscape of breakthroughs achieved in the recent 3D Computer Vision revolution

We believe that the exciting downstream possibilities are so darn lucrative that if not us, another startup filled with dreamers like us will win this vision. We are betting on us because we LOVE this and are fool-hardy enough to believe that this still matters.