#techploration

Slime molds are great at planning roads.

Know what that sounds an awful lot like? A neural network (or AI in marketing speak). Vast network of simple nodes that together complete complex tasks.

So what if you put slime mold up against ai?

What if the mold won? What if it turned out nature already perfected scalable neural networks and called them Fungus?

What if then people relied on the decisions of the fungus more and more

We can’t build a model that perfectly recreates the decision making patterns of the slime mold. It consistently beats our most advanced models (and massive simple models). There is something special with the mold

Something Divine

Questioning the decisions of the mold goes from unpopular to sacrilegious. The mold handlers are a religious political sect— the enforcers of the mold’s will.

Antibiotics are viewed as holy fire— penicillin is made from the blood of a god and has the power to kill a god

Add to the mix that an increasing amount of intellectual property is owned by private equity firms who have no interest in the content, just leveraging it for money

We are soon going to be back in the situation that led to Netflix’s initial success— huge media catalogs with zero distribution making zero money (once the short term tax write off passes)

This is beyond the pedantic ‘it’s not artificial intelligence it’s a neural network’ argument of terminology. This is purely branding.

AI is rapidly becoming a toxic word to have associated with your company. It has become shorthand for lazy and cheap. Three quarters of what is marketed as AI isn’t even AI.

We are at peak ‘shove AI in it’ right now as every app and piece of software is promoting their ‘AI enhancements’ and general privacy destroying bloatware. Give it six months before the combination of customer pushback and rising server/data costs means all these companies start quietly dropping their AI features

A year from now announcing the new ‘AI enhancements’ of your product will sound as ridiculous as saying you’re launching NFTs in 2024

Tech Bros are going to effectively kill ‘AI’ within the next two years. But what they are destroying the marketability of technology branded AI, not the actual power of neural networks.

The TV trope answer is rubber gloves and bleach. You do your best to not leave any, and destroy what is left. This is also the Gattaca approach, where in a society stratified by DNA you avoid detection of your ‘inferior’ DNA by … exfoliating?

This approach has obvious flaws. Imagining that you’ll never shed a hair or eyelash or slough off some skin cells seems naive. Encasing your body in a DNA encapsulating shroud is hardly subtle or convenient (hazmat chic has yet to happen). So what do you do?

You add some Noise

You know what is a lot easier than making sure no one ever gets a hold of your DNA? Making sure no one can reliably get a hold of your DNA, and that your DNA never shows up on its own.

You might think I’m going to suggest hacking your own DNA so that it changes constantly— let’s call that phase 2.

Phase 1– DNA spore cloud

Imagine this but all DNA

You set up simple bioreactor farms to churn out DNA strands (this tech already exists), get your sequence source from Ancestry.com and you get to work laundering DNA.

A misting of this DNA soup would make all other DNA functionally invisible— how are you going to separate out one set of DNA from the others?

And with the randomness of the seed data (and data breaches), your DNA being present has no legal bearing on you having been there in person.

The floating garbage patch in the Pacific Ocean a huge ongoing issue. Plastic is the worst offender (and no it’s not all drinking straws and plastic bags— it’s mostly discarded fishing nets). Skimming would be too costly and unrealistic (it’s country sized— big country). Plus is almost a biome at this point— you couldn’t scoop out the trash without also scooping animals/eggs/plants basically causing more havoc trying to clean it up.

So what do you turn to? Bacteria

So this area has been in active research for 25+ years as the ultimate solution to dealing with plastic waste.

Should be great right? The lifespan of a bottle in a landfill falls from centuries to weeks in a vat. It’s such an alluring goal that people gloss over the path

Brute forcing thousands upon thousands of mutations on an enzyme that specializes in breaking down hydrocarbons sounds wildly risky.

Because what if you get one that does too well?

Future Forecaste

Silicon Valley Tech Bro Billion wants to try and improve his image as a part of the private jet class by holding a big public competition for innovative solutions for the garbage patch. Encourages all of his tech bro buddies to pitch their ideas. Billionaire promises to fund a pilot project for the top idea.

What wins out? Bacteria

And they apply the Facebook ‘move fast and break things’ philosophy to brute forcing mutations. Garbage patch is in international waters, so no approval (or oversight) to go and test your ideas.

In fact, with being out in the middle of the ocean, you can build your lab right on a boat and sail out there. And test your iterations right there. In the ocean. Why test on a simulated garbage patch when the real one is right there?

Success! A strain that breaks down plastics in a short timeframe in the cold of the ocean! Your test site quickly goes from floating landfill to. something?

The enzymes broke the plastic into component nutrients. So you now have effectively dumped a whole flood of nutrients into the water. Kinda like dumping fertilizer

Cool. So now you you have turned garbage patch into the Great Pacific Algal Bloom and Dead Zone

And the bacteria isn’t staying put. Oceans have a way of moving things, and you just built a hardy sailor with a plentiful food supply.

The bacteria makes it to shore and suddenly, plastic isn’t permanent. Plastic isn’t safe in water. Every seal and gasket is now prone to failure. There is a rush to figure out which types of plastics are susceptible, which are resistant. New plastics with bacteria resistance are developed.

But that’s not even going to be the biggest issue

We’ve Got a Fuel Pox on our Hands

If it likes to eat plastic know what it’ll love? Gasoline and any other hydrocarbon

This bacteria would essentially turn gas into soy sauce. Think about fuel rotting

Suddenly world’s energy supply is at risk. Fear of contamination becomes the oil and gas industry’s number one concern. Gas becomes an even more precious commodity, and is only used when application demands. The industry takes on surgical level of cleanliness.

Not just assisted, but full autonomous driving. And it is going to fundamentally change our relationship with cars and transportation in general.

People are going to stop buying cars. The whole sales pitch of everyone buying their own self driving car is ludicrous— it fundamentally misses the opportunity that self driving actually presents.

Owning a car sucks. Having constant immediate access to transportation is a form of autonomy.

Both those things can be true. A car has to be maintained, insured, parked, replaced, protected, fueled— this whole laundry lists of responsibilities to maintain access to self directed transportation.

What about all the perks of having a car, but none of the hassle? That’s what a self driving car offers. A car when you need it, where you need it, without having to worry about everything else that goes along with owning a car.

Because you won’t own the car.

How much time do you actually spend driving? How much time is your car just sitting there? Why worry about and pay for a car you’re not driving?

Your car is going to be a subscription service

Uber is already testing this basic model, but in a world of self driving cars it makes perfect sense. You don’t own a car. You have a Car Subscription, which means there is a car there to drive you when you need— scheduled in advance or on demand. You pay for different subscription levels (pay per mile, unlimited, luxury, etc)

A personalized public transportation

People will realize owning a car is actually a burden, and a fleet of self driving cars that take themselves for servicing and refueling is actually a world easier.

There are going to be two major downsides

First, you are going to be tracked. Not just where you’re going but what you’re listening to and riding with on the way there. Think about it— you will not be able to anonymously go anywhere

Owning a car will become suspicious— an expensive luxury that offers anonymity. It will be like having a pager in the 90s— associated with doctors and drug dealers. Bikes and motorcycles will thrive in the ‘socially acceptable non tracked transportation’

Second major issue will be ads

The double edge sword of a self driving car is that it frees you up to do other things.

You think you are going to get to sit and enjoy life uninterrupted by ads during your morning commute? Your Hulu and Netflix are already synced— you buckle your seatbelt and your episode picks up where you left off. Spotify is connected. Your use profile instantly tailors the ride to your tastes

Just watch a couple ads first

You can always pay extra to go ad free. You’re just sitting there anyways. Also means they can finally get rid of billboards (or at least move them to inside the car). Short on funds? Watch ads your whole ride for a discount.

Large Language Models like ChatGPT are all the rage for investment right now, but 98% of the applications we see are complete misuses (see: Willy Wonka Experience).

So if you aren’t trying to save budget by firing writers (never going to work anyways), what’s an actual practical use for an LLM?

Translating internet speak

People who rely on screen readers complain that aesthetic typing quirks (homestuck/classic l337 spe4k) along with all the tiktok style censoring break readers. Screen readers aren’t anticipating numbers to sub for letters, or there to be a pronunciation for an emoji + letter contraction.

Don’t use the LLM to generate spam emails— have is read the post and translate all the internet speak. Have it read out ‘I have leet skills’ and not ‘I aich-four-vee-ee el-three-three-tee es-kay-one-el-el-ze’

It’s clear Matt fucked up, and there is a very real chance he won’t be able to come back from this without lasting changes. He isn’t a complete idiot, but also seems to have an ego as resilient as a damp tissue. Once he is back from sabbatical and his PR team has wrestled his phone out his hands, this is going to be the game plan.

The Scapegoat Buffer

The fact that he has gone into the DMs of random users to hash out this beef shows that he is not suitable for the role, and is likely going to be forced out of the position of Tumblr CEO.

But just Tumblr

He will remain the head of Automattic. Wordpress will still be his. He is not stepping away from his empire, only ceding control of a single fiefdom. But at the end of the day, he’s still king

Their pick will seem perfect at first. A progressive choice— a member of one of the communities targeted by Matt’s bigotry. A new CEO who Matt will be able to tout as a clear example of his learning and progress and blah blah blah PR bullshit because you know what?

They are going to be powerless.

Less than a figurehead. They are going to use them to try and cleanse their own image while slashing Tumblr’s budget even more. They will find a CEO that honestly wants change and progress and for Tumblr to live up the their ‘queerest place on the web’ badge of honor

And then they will give them none of the tools. No money, no resources (or rapidly dwindling resources), and absolutely. zero. support from above.

All So It Won’t be Matt’s Failure

He’s going to make sure of that. He can’t make Tumblr profitable. He doesn’t have the vision to move Tumblr into the future. So he is going to foist it on to someone else and make it their failure. He didn’t fuck up Tumblr, the next guy did. This probably sounds familiar because this is exactly what happened with Elon at Twitter. And it will be the same thing where the ‘Tumblr CEO’ will just absorb the blame while laundering Matt’s shitty ideas until Tumblr is finally pulled from life support