10 Predictions from the year 2016 and how they panned out

In 2016 I sat down with friends and cousins and we all made predictions about what would take place by the end of 2026. We’re close enough to 2026 that I thought it would be fun to go through the predictions and see what (if anything) we got right and where we were off base. I’m going to rate the predictions on a scale of 0-10, with [0] being a total miss and [10] being spot-on. We made more than 20 - but I’m just going to post 10 here for now.

Prediction #1: E-butlers will exist that can make small decisions on our behalf.

Rating: 8

We’ve got 28 months to go and given the explosion of interest in Agents (see many of my previous posts including the most recent one), this seems likely. My favorite post on the topic is this one - published in January 2023:

Prediction #2: Drone delivery of virtually everything

Rating: 0

We were only referring to light-weight objects. Still, obviously this hasn’t happened. UPS released a drone-delivery teaser 7 years ago (never went anywhere):

We’ve also watched Amazon spend $100 Billion+ building delivery infrastructure that has nothing to do with drones - though it has succeeded in becoming the largest shipper after USPS (note chart below ends in 2021).

As of 2023 Amazon parcel volume was 20% ahead of UPS and the gap is growing quickly. It is tracking to overtake USPS in the not distant future.

Given how long it would take to build out drone infrastructure the 2026 prediction is a guaranteed complete miss. Recently investment into drone delivery has increased (see below), but I don’t think the statement “of virtually everything” will be true - ever.

Prediction 3: Virtually no crime outside of inner cities

The thought here was that technology would make it easier to catch criminals - hence criminals would be less likely to commit crimes in the first place.

Rating: 0

While crime rates continue to fall with the exception of large cities, they’re nowhere near “no-crime in the suburbs”:

Prediction 4: All vehicles are driverless

Rating: 0

While FSD 12.5 is impressive, even if it were to become perfect tomorrow it would take far more than 28 months for all vehicles to go driverless. My current expectation (which changes on the daily) is that level 5 autonomy might not be solvable without essentially solving AGI - which would put the timeline around 2028-2032 (my current guess).

Prediction 5: Goods smaller than 3’x3’ with only a few raw material inputs will be available as software downloads for 3d printing at home

Rating: 0

I’m going through these predictions in real-time - I didn’t bother reading the list in advance. I’m noticing a pattern - which is that we were all caught up in believing that concepts featured in Ted-talks and futuristic videos would simply come to fruition at scale. “Look Ma! I see it right there, obviously we’ll have it within 10 years!”

I’ll say something else - we’re obviously a group of pathologically optimistic people, and I feel like that’s a good thing!

The most impressive capex projects in all of history (outside of China) have been Tesla’s production ramp and the build-out of Amazon’s fulfillment network. I’ve watched each over the past five years (and have also been running a manufacturing business) - collectively these and other similar observations and experiences have given me new appreciation for how long it takes to build and adopt new processes.

My current view of 3d printing is that despite improvements in resolution and material science - it won’t make economic sense to “print” anything that is mass produced until we have nanobots - and at that point the process won’t be described as “printing” anyway.

Prediction 6: Cancer will be like diabetes is today

What we meant by “like diabetes” is that it is manageable and most people can have a near-normal life expectancy with appropriate treatment.

Rating: 0

Cancer cases and death rates continue to fall:

Still, deaths from diabetes hover just north of 100,000 - while cancer remains above 600,000. The below table shows leading causes of death in 2022:

Prediction 7: All new vehicles are electric

Rating: 0

Still, the growth of the EV market over the past 8 years has been remarkable, and there’s reason to believe we’ll see an acceleration in EV sales around 2 years from now after the pause we’re experiencing. Here’s some data.

Almost all cars sold in Norway are EVs. More than 40% of vehicle sales in China this year will be EVs. I found a super interesting report on EV trends written by the IEA (linked here) - the opening paragraph is printed below:

Almost 14 million new electric cars1 were registered globally in 2023, bringing their total number on the roads to 40 million, closely tracking the sales forecast from the 2023 edition of the Global EV Outlook (GEVO-2023). Electric car sales in 2023 were 3.5 million higher than in 2022, a 35% year-on-year increase. This is more than six times higher than in 2018, just 5 years earlier. In 2023, there were over 250 000 new registrations per week, which is more than the annual total in 2013, ten years earlier. Electric cars accounted for around 18% of all cars sold in 2023, up from 14% in 2022 and only 2% 5 years earlier, in 2018. These trends indicate that growth remains robust as electric car markets mature. Battery electric cars accounted for 70% of the electric car stock in 2023.

Global EV car stock is still going parabolic thanks to China:

And here’s EV sales by region - China is dominating (orange):

There’s a big difference between the manufacturing processes and supply chain used to make EVs and that used to make ICE vehicles (internal combustion engines). EV bill of materials and total cost of goods are still falling rapidly as the production of core components (especially batteries) and raw materials expands. It is a certainty that prices will continue to fall to the point at which ICE vehicles will no longer be cost competitive without government intervention. BYD introduced the “Seagull” in 2023 and is selling 35,000+ per month, with 350,000+ units sold to date:

In China it sells for $10-12k, it is already selling in Brazil and other South American markets, and it is expected to launch in the UK for <$19k.

I also ran across an interesting Chinese company selling EVs for less than $5k on Alibaba, like this one for $2,200:

Here are it’s stats:

This car has electric scooter like speeds, but can drive up to 80 miles on a single charge, and can carry a load of 1100 lbs. The reviews on these vehicles seem pretty good based on my two minutes of research. Batteries account for up to 40% of the cost of an EV, and prices will have fallen by about 40% from 2022-2025 according to Goldman Sachs:

It’s actually remarkable how quickly EVs have grown given their historically higher cost, lack of consumer awareness about current costs, lack of public charging infrastructure, and up front cost of in-home charging infrastructure. Given the rapid price declines it seems likely that EVs will become the dominant mode of transportation with the potential (likely even?) exception of hauling heavy materials over longer distances.

Prediction 8: Parabiosis will be available as a longevity treatment

Parabiosis is a surgical technique where two animals (or humans) are joined in such a way that they develop a shared circulatory system, allowing for the exchange of blood and other bodily fluids. It seems unlikely that transferring plasma/blood from a young person to an older person will be effective in improving health. Bryan Johnson - the “most measured man on the planet” is currently spending $2m+ per year trying to slow down his rate of aging. He is a fascinating person to follow and I highly recommend doing so. He experimented with plasma transfusions (his son providing him plasma, and he providing his 70 year old father plasma). The results indicated there was no benefit so he discontinued the exchange (link to source here).

I’ve always been skeptical about the efficacy of simply transferring blood and plasma - but what makes a lot more sense is full blown parabiosis. Why? Because in a parabiotic state the younger person is processing the older persons blood:

In experimental setting, researches have observed that younger animals in parabiosis experiments can influence the physiology and regenerative capacity of older animals. This includes improved tissue repair and possibly even the rejuvenation of certain tissues, suggesting that factors from the younger animal’s blood, including stem cells or signaling molecules, may contribute to this effect.

Promising experiments have been done with mice (link to source).

Intuitively, parabiosis seems like it should work. The ethical question that needs answering before people start experimenting is will there be any lasting damage to the young person?

Part of this prediction is predicated on young people being willing to take the risk in return for a substantial sum of money (you need to be hooked up for many hours - so the only way this happens would be if ultra-wealthy individuals pay enormous sums of money). Given the rate of advancements in medical technology, it might very well be worth it to a person who is 22 years old to make $250,000 a year in return for potentially accelerating their own aging - given that they would still have many years for medical treatments to advance and repair any damage they incur.

Unfortunately, I have to give this prediction yet another: 0. There is no indication that humans are working on setting up these experiments.

Prediction 9: Refined wetware interfaces will be available for the disabled

WAHOOOOOOOOOOOOOOOOOOO! For the first time in what feels like ages, we don’t have a 0 prediction score. I’m going to give this one a ‘3’. Our expectation was that these interfaces would be available as an actual product on its way to becoming a standard of care. That hasn’t happened. However, thanks to the advancements made by Elon Musk’s Neuralink I think it’s fair to give us more than a zero - especially considering we still have 28 months to go.

Wetware refers to technology that directly interfaces with the brain (the “wet” part). While Neuralink was not the first company to install an implant in a brain capable of moving a cursor, it IS the first company to achieve many milestones, including:

1. Number of electrodes connected

2. Minimizing invasiveness of surgery via advanced robotics

3. Speed of communication between the brain and computer input is vastly faster - enabling a far superior user experience (ability to play Mario Kart, for example)

Links to more info:

1. Behind Elon Musk’s Brain Chip

2. Demo of Neuralink Chip in Quadriplegic Noland Arbaugh

Prediction 10: Voice interaction with most IOT objects

Rating: 2

The idea here was that any object connected to the internet would offer voice as an interface - replacing or augmenting buttons and touchscreens. I think this deserves more than a zero because of the advancements we’ve seen with generative AI and LLMs.

Until recently using voice as a “Fourth Component of the GUI” (see below post)

was completely infeasible because machines couldn’t decipher most of what we said to them. Alexa, Hey Google, Cortana, Siri (etc) were all a total joke that hadn’t seemed to improve in the past ten years. LLMs changed that completely. Now LLMs can not only understand us perfectly - they can also infer what we’re trying to say even if we jumble our words up.

Now that we’ve cracked the tough part (teaching machines to understand language) it is only a matter of time before voice interfaces become a core component of how we communicate with technology.