"Six thousand years ago, someone invented the plow, and we all got wealthier. A gentle reminder that all civilizational wealth is driven by invention." - X bio of @JeffBezos
173
Arun Shroff retweeted
If an alien visited earth and asked you: “Show me one chart that sits upstream of everything your species has accomplished or will ever accomplish.”
What would you show them?
I would show them the chart below.
From my perspective, as long as we keep this chart going, we will figure it all out…
I discovered Penrose's arguments about consciousness not being computable, many years back as well.
teapaw.org/arunshroff/status/1607…
Penrose's argument, in a nutshell, is that since humans can understand Goedel's theorem, consciousness is uncomputable and physics incomplete. I've always found it quite charming, the idea that understanding Goedel's theorem is an indicator for consciousness, seeing how few people actually understand Goedel's theorem.
In any case, the argument from Goedel's theorem is why, I believe, Penrose thinks that the measurement process in quantum mechanics is relevant for human consciousness. Because it's the only place in the foundations where you can fumble in some new physics that is plausibly relevant in the human brain. Hence the story with the microtubules. And if you believe that, then AI can't be conscious because it's algorithmic. Though presumably a quantum computer could be conscious, if it emuluates the quantum process that is supposedly relevant for consciousness.
(Penrose's model for wavefunction collapse by gravity is totally computable and actually does not fit that narrative.)
Also, I believe Penrose's argument from Goedel's theorem to be technically wrong. It is possible to algorithmically understand (and even prove) Goedel's theorem without violating Goedel's theorem because the theorem itself is a meta-statement about the Goedel sentence(s) and not the sentence itself. In some sense, Penrose's argument is wrong for the same reason Goedel's theorem exists in the first place, that you can make meta-statements about statements. In fact there have been computer-generated symbolic proofs of Goedel's theorem long before the AI boom.
It's not that I have an obsession with Penrose in particular, it just so happens that my interests have a big overlap with his. I have an interview with Penrose in my 2022 book "Existential Physics" on these and related problems.
teapaw.org/realBigBrainAI/status/…
3
162
Arun Shroff retweeted
Prediction: In ~ 5 years or less, most of us will be using a chatbot like this, to answer our questions rather than Google search.
This is the biggest threat to Google. Search will still be used but for different things. Google is probably already working on a similar chatbot.
Arun Shroff retweeted
Lost my phone at the office and spent 30 minutes turning the place over. Find My was disabled by MDM.
Out of ideas, I asked Claude how I could find it. It suggested tracking the Bluetooth signal strength, then wrote me a meter in about a minute.
I walked around watching the number climb. Found it.
Apparently you can just make the tool you need now.
Code: github.com/ben-z/findphone
Arun Shroff retweeted
GPT-5.6 Sol for resolving 100+ year old conjectures.
Wild that this level of intelligence can be accessed by and is available to empower everyone!
The Maxwell conjecture is false. Counterexample found by AI (GPT-5.6 Sol), communicated by human mathematicians: arxiv.org/abs/2607.27197
Arun Shroff retweeted
Catching skin cancer early is a home robotics problem.
Melanoma is highly treatable when detected early, yet today’s screening process depends heavily on patients noticing tiny changes across their entire skin surface. This requires patients to solve a near-impossible visual-memory and registration problem.
I built OpenDerm, an open-source 4-DOF robot that captures high-resolution images of the skin and uses them to reconstruct and track the skin surface in 3D over time.
The best way to make skin screening truly routine is to bring it into the home. OpenDerm shows that inexpensive robotic skin imaging is possible, but the path to scale is not a dedicated screening robot in every household—it is to make skin screening one of the many useful things a general-purpose home robot can do.
Read more about why I built OpenDerm and how it works here:
Blog: marionlepert.github.io/blog/…
Project: openderm.github.io/
This seems quite prescient now...
Software is eating the world, was true till recently. Now AI is eating software.
I do believe this is the future of software development. In <5 years, AI powered assistants like this will replace low-code tools to create complete working applications based on user input.
1
1
136
Arun Shroff retweeted
A prediction from the past that I got right
Replying to @othman
"What ChatGPT in particular has made us realize is that many of the tasks that we have so far labeled "search" and where we click on blue links are really tasks of research, analysis and decision-making."
For what it is worth, I was suggesting this all the way back in 2015 - that if Microsoft wanted to disrupt Google, the best way to do it would be to answer questions rather than provide links to search results. It would be the opposite of a search engine - a "find" engine.
And although we did not have LLMs or ChatGPT back then, the closest thing to conversational AI we had then was Wolfram Alpha, which answered questions in natural language. I had suggested that Microsoft acquire them, and integrate the technology for answering questions into Bing. Here is the thread from back in 2015:
teapaw.org/arunshroff/status/6795…
Arun Shroff retweeted
I solved 6 open Erdős problems in 5 days, using @OpenAI GPT-5.6 Sol.
I have a math background, but the Codex workflow I used does not require deep mathematical knowledge.
Here’s exactly how I approached it, including my prompts 🧵
This is pretty remarkable! Wonder how long before we can do this in other disciplines - like physics, medicine etc.
@ericweinstein - maybe this is how we solve the current bottleneck in physics progress, and figure out how to travel to the stars...
I solved 6 open Erdős problems in 5 days, using @OpenAI GPT-5.6 Sol.
I have a math background, but the Codex workflow I used does not require deep mathematical knowledge.
Here’s exactly how I approached it, including my prompts 🧵
1
99
Arun Shroff retweeted
In this piece I pick the simplest term of the standard model, and explain it piece by piece, so that the rest of the terms make sense. I think that's a useful way of going about learning mostly anything complicated. A great question to ask yourself in philosophy / math / physics is, “what is the smallest amount that you need to know such that when you understand this one case, you understand the majority of the rest?” In other words, for which relatively elementary case do you get the largest bang for your comprehension-buck?
LINK IN REPLY
Arun Shroff retweeted
Feels like a watershed moment for advancing mathematics. Scientific and medical advances which can really improve people’s lives feeling very close now.
So far GPT 5.6 Sol is the most interesting new model mathematics-wise for me - those of the new proof claims on erdosproblems.com that I've looked at in detail have all been correct, and moreover contained some interesting ideas.
1/7
Arun Shroff retweeted
Engineers at Coinbase are getting closer to recursive self-improvement, or loops.
For instance, agents can now ingest and summarize customer feedback collected right in the app, prioritize bugs and features based on that feedback, draft the code, security review it, and provide it to a human engineer for final review. All happens automatically each day and it learns from the human edits to improve over time.
We're finding more areas to introduce loops across the company. Instead of prompting agents with what to do, you can increasingly give them a goal, and they bring back high quality work for you to review.
The fundamentals of engineering haven't changed, but the daily work has.
We're managing agents, building loops and taste and judgment matter more than ever.
So we rebuilt how we hire engineers from scratch to match how we build. Here's what we've been up to @ Coinbase:
coinbase.com/blog/interviewi…
Andrej Karpathy just dropped a 6-hour course on how to build LLMs from scratch:
• 00:00 - Deep dive into LLMs like ChatGPT
• 03:31:23 - Building ChatGPT from scratch in live
• 05:27:43 - How to use LLMs (Karpathy method)
This course will replace a $90K Stanford LLM master’s degree.
Start watching today, then read how to become an AI engineer in article below.
Arun Shroff retweeted
Introducing NEO’s 25 Degrees of Freedom, tendon-driven hands — nearing or surpassing human-level dexterity, strength, speed, and reliability.
For seventy years, robotics worked around the hand problem. The humanoid bet is the reverse: it lives or dies at the fingertips.
Arun Shroff retweeted
I did the math, and $2M in OpenAI tokens is equivalent to a 450 pre-AI human engineering team for 2 years, in terms of raw LOC output.
OpenAI is offering $2M in tokens to every YC company in the spring and summer batches.
We extended the summer deadline to May 25 so more founders can get in on it.
ycombinator.com/apply
Arun Shroff retweeted
Uber CEO Dara Khosrowshahi said earlier that currently, 90% of Uber’s engineers use AI, but the top 30% (power users) are seeing unprecedented productivity gains.
These power-users of AI are pushing the maximum number of "diffs" to the codebase.
He predicts in 5 Years the ROI of a human engineer is surpassed by the ROI of adding more AI agents and GPU power. So at that time he will just hire more AI agents and pay for NVIDIA GPUs instead of human software engineers.
---
From 'The Diary Of A CEO' YT Channel (link in comment)
A prediction from the past that I got right
Replying to @othman
"What ChatGPT in particular has made us realize is that many of the tasks that we have so far labeled "search" and where we click on blue links are really tasks of research, analysis and decision-making."
For what it is worth, I was suggesting this all the way back in 2015 - that if Microsoft wanted to disrupt Google, the best way to do it would be to answer questions rather than provide links to search results. It would be the opposite of a search engine - a "find" engine.
And although we did not have LLMs or ChatGPT back then, the closest thing to conversational AI we had then was Wolfram Alpha, which answered questions in natural language. I had suggested that Microsoft acquire them, and integrate the technology for answering questions into Bing. Here is the thread from back in 2015:
teapaw.org/arunshroff/status/6795…

























