14 months apart
20 Jan 2025
“They were able to suck up all the text on the internet and use that text to make a prediction on what the answer should be when someone asks a question word by word.”
spoken ·Seed World Group·source
27 Mar 2026
“There's never been an empire that's been in the late stage that suddenly got a technology like AI.”
spoken ·Suffiyan Malik·source
Friedberg proposes 3D printing, biomanufacturing, and automation could enable US to decouple from Chinese production through reinvented manufacturing methods.
“And so through automation, three d printing, through biomanufacturing, you can kind of change the methods of doing this stuff.”
Friedberg argues sustainability comes from technology enabling more consumption, not convincing consumers to consume less.
“I'm a big believer that sustainability in the twenty first century does not arise from convincing consumers to consume less. I think sustainability arises from building technology based solutions that let consumers consume more”
Friedberg argues sustainability comes from technology that enables more consumption at lower environmental impact, not from consuming less.
“I think sustainability arises from building technology based solutions that let consumers consume more and dropping the price and dropping the environmental impact.”
Friedberg argues sustainability should focus on increasing consumption through productivity gains rather than reducing consumption.
“Can we give people 10 times, 100 times, 1,000 times more than they have today? And in doing so, it makes us solve the productivity problems. And the productivity problems ultimately resolve the sustainability problems,”
Friedberg notes DNA sequencing costs have fallen faster than Moore's Law over twenty years.
“Yep. Now digital technologies have driven the cost down of DNA sequencing faster than Moore's Law in the past twenty years.”
Friedberg describes a system using seven to nine plant enzymes in tanks to make starch from atmospheric carbon.
“And you put them together in a bunch of tanks, and then they can take carbon out of the atmosphere and they can actually make starch.”
“At that time, probably the largest we were known to be the largest users of Amazon's Elastic Compute Cloud because we were ingesting so much data to build these models that related all the variables of farming and the weather to predict outcome.”
Friedberg contrasts 1950s technology optimism with current negative associations toward technological progress.
“Go back to the 1950s. There were all of these crazy posters about humans will go on rockets to the moon and we're all going to live on star bases on the moon.”
Friedberg explains that in the 1970s Disneyland remade Tomorrowland rides to reflect fear of technology going wrong.
“In the nineteen seventies, they started to turn over all the rides in Tomorrowland at Disneyland and every ride got remade as a fear of technology ride.”
Friedberg says one gene therapy death led to a seven-year halt despite dozens of life-saving therapies now available.
“They weren't allowed to do any more gene therapy treatments. And since then we now have literally dozens of gene therapy treatments that can cure dozens of human diseases saving millions of lives.”
Friedberg notes that Disneyland's Tomorrowland in 1955 originally celebrated future technology opportunities optimistically.
“In 1955, Disneyland opened, I've told this story before with this ride called Tomorrowland. This area called Tomorrowland. I mean, you've been there. And every ride there was about the future opportunity with technology.”
Friedberg developed boosted breeding to enable true seed propagation in crops currently using vegetative propagation.
“We've developed a new system of of breeding that we call boosted breeding that really changes how we do breeding, the pace of gains, and also allows for true seed propagation in plants where we're typically doing vegetative propagation where there isn't a seed business today.”
Friedberg explains boosted breeding creates tetraploid true seed in potato, enabling higher yields through genetic uniformity.
“In our system, we can actually combine two hybrid or two kind of genetically distinct diploid potato to make a tetraploid offspring, but all the seed is tetraploid.”
Friedberg explains LLMs are text predictors using internet data to generate word-by-word predictions, not true AI.
“They were able to suck up all the text on the internet and use that text to make a prediction on what the answer should be when someone asks a question word by word.”
Friedberg paid $10,000 for 17 megabytes of weather data in 2006, now equivalent to two iPhone photos.
“It was 17 megabytes of data. That's like as much data as you get when you take a photo on your iPhone now or two photos on your iPhone pretty much.”
Friedberg reports new German AI architecture achieved 7000x speedup over Jetson Nano and 100x over H100.
“Their architecture led to a speed up of 7,000 x compared to the NVIDIA Jetson Nano, 300 x compared with NVIDIA RTX forty ninety, and then a 100 x compared to the NVIDIA h 100.”
Friedberg was the first person to review Final Frontier's app while still a student preparing for exams.
“Like, I was, I believe, the first person to review their, their app and a few other things and give feedback on it.”
Friedberg says AI researchers are surprised that making output recursive drives major model performance improvements.
“I saw a lot of computer scientists that have worked in AI for some time, I think, be a little bit surprised about this moment that we're in, that we're seeing such incredible strides in model performance just by making the output recursive.”
Friedberg argues no late-stage empire has ever gained a technology like AI before.
“There's never been an empire that's been in the late stage that suddenly got a technology like AI.”
Friedberg explains material would be moved off the moon using electric rail guns rather than propulsion systems.
“But the mechanism for moving material off the moon is not propul propulsion. It's actually a mass driver, so electric rail gun.”
Friedberg says federal government spent hundreds of millions on useless workforce training for desktop computers in the eighties.
“In fact, the federal government did this deal in the early eighties. They spent hundreds of millions of dollars in workforce training programs that literally did nothing.”
Friedberg categorically rejects the idea that AI will destroy jobs, calling it a Luddite narrative disproven daily by what he observes on the ground.
“There is no job loss with AI. I’ll say it again. I’ve said it a thousand times, and I’ll say it again and again and again. The idea that AI is going to destroy jobs is a Luddite idea that is being disproven every single day. I see it on the ground. It is only a matter of time before people wake up to this and realize that this narrative that they’ve been sold is a crock of bleep.”
Friedberg describes a medical intervention as being more accurately classified as a technique rather than a conventional pharmaceutical.
“It is much more of a technique than a drug in the traditional sense.”