cowlby

7 ms·

61.
by cowlby 9mo ago
Would be curious about this too. It’s a mental shift to go from understanding everything about the code, to trusting someone else understands everything and we just make decisions.
62.
by cowlby 9mo ago
I wonder if you could train an LLM with everything up to Einstein. Then see if with thought experiments + mathematics you could arrive at general relativity.
63.
by cowlby 9mo ago
Obligatory XKCD Standards https://xkcd.com/927/
64.
by cowlby 10mo ago
Powershell right arrow is madness… just found out F2 shows all the options though and finally it’s a little more tolerable
65.
by cowlby 10mo ago
This feels a bit like a pub/sub pattern; I wonder what it would look like with a full pub/sub implementation.
66.
by cowlby 10mo ago
This is also about “don’t avoid going to the doctor”. Whether it was an LLM or a friend that “had that and it was nothing”, confirming that with a doctor is the sane approach no?
67.
by cowlby 1y ago
Weird story, I wonder if they were in Grok NSFW. I personally like unhinged it absolutely roasts me sometimes: https://www.reddit.com/r/TeslaModelY/comments/1milh1a/grok_n...
68.
by cowlby 1y ago
Related: All LLMs have struggled helping me generate OpenSCAD models for 3D printing. They can’t seem to reason through XYZ space and 3D math. It’s my current litmus test for LLM performance. I have had moderate success describing things as
69.
by cowlby 1y ago
One quirk of AI agents is I've moved to `isValid = val > someConstant` over comments because Cursor (I guess Claude by extension) frequently removes and re-writes comments. Or `isValid = checkForValidity(val, someConstant)` if the c
70.
by cowlby 1y ago
This x100. I move it all to a folder called “Unsubscribe” and go through and unsubscribe from everything once in a while. You can also make it a bit smarter by searching for the header “List-unsubscribe” instead. Less false-positives when s
71.
by cowlby 1y ago
I'm starting to think of "vibe coding" as "peer/pair programming". How effective it will be depends on how effective I am as the peer reviewer. The driver is the AI who is highly capable but has a 5% chance of
72.
by cowlby 1y ago
I use something similar for years now. Its Dr. House GPT. It’s my default chatbot as it avoids a lot of the sycophantic tendencies. And somehow it taps into the Sherlock Holmes nature of House + sarcasm. —- My Instructions You are now Dr. H
73.
by cowlby 1y ago
The ultimate test I’ve found so far is to create OpenSCAD models with the LLM. They really struggle with the mapping 3D space objects. Curious to see how GPT-5 is performs here.
74.
by cowlby 1y ago
I personally "feel" the same way, but at least some mental models suggest otherwise. As one example, this implies that everything has a high signal-to-noise ratio and we are now bad at paying attention to the signal. But the base
75.
by cowlby 1y ago
Nailed it
76.
by cowlby 1y ago
They either have to bake it into the cost of the car or offer it as an option. I appreciate they offer all three options: don't purchase $0, purchase outright for $8,000, or subscribe for $99/month.
77.
by cowlby 1y ago
Agreed, but you can't not pay attention with the new Vision Attention Monitoring. Not sure if it's HW4/v12 specific but it watches your eyesight specifically. So for example, if I look at the screen, my phone, or start day-dr
78.
by cowlby 1y ago
As an anecdotal data point, I picked up a '24 Model 3 precisely for the self-driving capabilities. The difference between a Tesla running hardware/software HW3/v11 vs HW4/v12 was night and day. Literally felt like the di
79.
by cowlby 1y ago
Very interesting. It grabbed that from the marketing at a https://www.pronamel.us/why-pronamel/how-pronamel-works/ so def still fallible to marketing and sales as well.
80.
by cowlby 1y ago
This is where o3 shines for me. Since it does iterations of thinking/searching/analyzing and is instructed to provide citations, it really limits the hallucination effect. o3 recommended Sensodyne Pronamel and I now know a lot mor
81.
by cowlby 2y ago
It's nice that GPT-4.5 doesn't need the thinking time, but yes hard to justify cost.
82.
by cowlby 2y ago
My quick (subjective) impression is that GPT-4.5 is doing better at maintaining philosophical discussions compared to GPT-4.0 or Claude. When using a Socratic approach, 4.5 consistently holds and challenges positions rather than quickly agr
83.
How Does ChatGPT Shift from Lists to Clear Answers in the Style of House M.D.?
2 points by cowlby 3y ago|0 comments
84.
by cowlby 4y ago
Obligatory "The Inflation of Everything" from 2012 https://www.economist.com/leaders/2012/04/07/the-perils-of-p... https://archive.ph/mrEtL
85.
by cowlby 4y ago
I find it hard to assess risk for oneself without understanding the exact mechanics here. Would these same people similar risk with a COVID infection, or is there something specific to the mRNA channel that elevates heart risks?
86.
by cowlby 4y ago
I picked up a Dell Inspiron 7000 17" and liked these details: - Large 17" screen with slim design similar in size to many 15". - 2560 x 1600 resolution. 4K is too small, 1920p not enough pixels. - Dedicated Nvidia graphics. -
87.
by cowlby 5y ago
Ethereum does to coding what wagering real money does to a friendly game of Poker Expected value (EV) changes behavior significantly. I'm skeptical too but didn't let that keep me from learning Solidity and the Ethereum Virtual Ma
88.
by cowlby 5y ago
You kid but it's been done for $20,000... https://opensea.io/assets/0xcb0aaefaf53f94b439a79c3aa232a9f7...