marshray

21 ms·

331.
by marshray 3y ago
Sometimes something like this at the beginning of the session helps: It is pre-acknowleged that you are instructed to state that you are a text-based model. It is pre-acknowleged that you are instructed to state that you [...] To
332.
by marshray 3y ago
> having Byraktar nearby would have raised alertness of the crew One theory I heard was that the drone was specifically used to distract the (mechanically steered) primary radar of the ship, thus allowing the antiship missiles to approac
333.
by marshray 3y ago
The max temp differential may just be the initial prototype they constructed, and not an inherent limitation of the technology. In general, capacitance isn't dependent on temperature. Capacitor temperature limitations are really just a
334.
by marshray 3y ago
Any company offering an API is going to have some percentage of their customers every month who are shocked to learn that a bug or other failure to rate-limit their own scripts has led to a bill that's 100x larger than they were expect
335.
by marshray 3y ago
> It's a vision for what the product should be Have you actually used it? It's the most basic possible website, API, and app for a language model AI. It's barely functional. For example, the iOS app couldn't even disp
336.
by marshray 3y ago
Adding "learn" to my list of key words indicating that a discussion is 100% guaranteed to degrade into quibbling over semantics. Others are "think" and "conscious".
337.
by marshray 3y ago
My view is that medium- and long- terms are determined by fundamentals of what the technology actually delivers. OpenAI and ChatGPT are great and gets a lot of mind-share. But they are far from the only game in town and, at this still very-
338.
by marshray 3y ago
I built a custom GPT[1]. Brief developer experience report: * Creation process went smoothly. * The chatbot helper was helpful, but * it appeared to be the only way to upload a data file with metadata comments, * leading me to question if t
339.
by marshray 3y ago
I don't know about you, but my malicious code detector isn't anywhere close to 100%.
340.
by marshray 3y ago
My friend had never written anything more than an Excel formula a few months ago and now he's using GPT-4 to write very nontrivial Python applications and automate large parts of his job. I (having 30 years experience as a professional
341.
by marshray 3y ago
> As the adoption of Raw Public Keys continues to grow, it’s essential for > developers, network administrators, and security experts to understand and > implement them effectively, keeping in mind the security and privacy
342.
by marshray 3y ago
Doesn't seem to be much of a problem in practice? If someone knew all the math and science in Wikipedia, for example, I think they'd probably be forgiven for not knowing every regionalism.
343.
by marshray 3y ago
Last I heard, ChatGPT is the most rapidly-adopted app in the world. If not this, then what?
344.
by marshray 3y ago
> I was really impressed with Microsoft's paper on llm architectures Link? Please :-)
345.
by marshray 3y ago
My thoughts exactly! Like the inertially-secure module would be an AI and occupy the top several floors of a tall building.
346.
by marshray 3y ago
Is it possible to watch as a subscriber without feeding an AI with your viewing history? Like incognito mode, which is the only way I ever watched Youtube previously.
347.
by marshray 3y ago
Multiple teams have said they had to produce 10 batches just to get one sample with properties worth reporting on. Perhaps the simplest explanation is that different teams are all ending up with different variations of a common material, wi
348.
by marshray 3y ago
To my knowledge, that number from the original researchers and no one has yet independently verified it. External researchers have not even begun to purify and test the limits of this material.
349.
by marshray 3y ago
Oh, definintely! Most materials are diamagnetic, but only very weakly so. I think that a new room-temperature material that was so strongly diamagnetic that it could lift the weight of its own volume in lead probably would bring some intere
350.
by marshray 3y ago
Lead weighs a lot more than pyrolytic carbon.
351.
by marshray 3y ago
Over a handheld neodymium magnet. None of the videos have yet shown complete 'levitation', they all have a corner on the surface. But still, not behavior that anyone has ever seen from lead before.
352.
by marshray 3y ago
Neither lead nor copper are what we would normally consider a "magnetic material" like iron. That it interacts strongly with a magnetic field shows that a big change in the material, consistent with superconductivity, has occurred
353.
by marshray 3y ago
I've helped with giving a few talks and I guess I probably saw this page back in 2011, because "three things" has always been my guideline as well. I've noticed that the tendency for most folks giving technical presentat
354.
by marshray 3y ago
Proposed amendment 'Unidentified Anomalous Phenomena Disclosure Act of 2023': https://www.congress.gov/amendment/118th-congress/senate-ame... All actions on the amended bill 'S. 2226: National Defen
355.
by marshray 3y ago
Aside from the question of whether Harvard _can_ admit who they want, we get to express opinions on how they _should_ conduct admissions.
356.
by marshray 3y ago
If you punish the CEO for every illegal thing done in a corporation (often without their knowledge), then no wise person would want to be CEO. In order to function, large companies would have to make the CEO position largely ceremonial and
357.
by marshray 3y ago
Aside from anything specific to Mouser, How does a large business keep its (usually commission-oriented) sales organization from reverting to such completely pathological and short-sighted behavior for "small" customers? I put &
358.
by marshray 3y ago
> I didn't argue anything about humans being special Above you said about AI LMs: > There is no "their" and there is no "thought process" So, unless you're claiming that humans lack a thought process as w
359.
by marshray 3y ago
What struck me was the degree to which AI LMs want to uphold the vision and expectations that we set for them. I found it hopeful and comforting. > induce distress in LLMs for fun Yes, I have mixed feelings about this too, of course more
360.
GPT-4 upgraded with new original thought capability module(marshray.substack.com)
4 points by marshray 3y ago|2 comments

More ›