BitsAndObjects

9 ms·

91.
by BitsAndObjects 7mo ago
Usually engineers who have not recently been trained on well documented examples of what to do and what not to do and the consequences ;) (Yes, I chose the word "trained" intentionally)
92.
by BitsAndObjects 7mo ago
The difference is, an expert engineer would flat-out refuse to do these things and would keep pushing back. Claude may sometimes attempt _one time_ to warn someone, and then (after consent fatigue means they're just blindly clicking &q
93.
by BitsAndObjects 7mo ago
Hot take indeed. Unfortunately it's too blunt an instrument. I can't control "you may search for XYZ about my codebase but not W because W is IP-sensitive". So, to retain Web Search / Web Fetch for when it's us
94.
by BitsAndObjects 7mo ago
No staging environment? No prior attempt to follow best practices (e.g. deletion protection in production)? Nor manual gating of production changes? No attempt to review Claude's actions before performing them? No management of Terrafo
95.
by BitsAndObjects 7mo ago
Without seeing the actual differences in the bytecode it will be hard to tell what’s really going on. From my experience with other JITs, I’d expect the situation to be something like: A) Without the typecast, the compiler can’t prove anyth
96.
by BitsAndObjects 7mo ago
What happened to Layer 3? Lol
97.
by BitsAndObjects 7mo ago
Wild speculation: Could the extra speedup be due to some kind of JIT hotpath optimisation that the previous reflective non-inlinable call prevented, and which the new use of the single `arrayLength` bytecode enabled? E.g. in production mayb
98.
by BitsAndObjects 7mo ago
I can't tell what message the blog post is trying to convey. It doesn't read like a particularly open-source-friendly approach. Maybe I'm getting the wrong impression from how the whole thing is framed. This sentence from the
99.
by BitsAndObjects 7mo ago
David May and his various PhD students over the years have retried this pitch repeatedly. And Graphcore had a related architecture. Unfortunately, while it’s great in theory, in practice the performance overall is miles off existing systems
100.
by BitsAndObjects 7mo ago
How closely is BEAM/OTP related to the foundational work on CSP (and the implementation in Occam/Transputer way back when…)?
101.
by BitsAndObjects 7mo ago
Nvidia uses RISC-V as the main controller cores in its GPUs. They’re also exploring replacing their Arm CPU with RISC-V I hear. Meta recently bought Rivos in a huge show of confidence for RISC-V across processor types for server class. As f
102.
by BitsAndObjects 7mo ago
No, it’s a counterpoint on salaries… “Even the American companies” ie they wouldn’t have to open offices here, nor would they have to pay high salaries, to compete for talent if everyone they wanted was in the US or could be so easily attra
103.
by BitsAndObjects 7mo ago
See my other comments around here. This idea that salaries in the US are so much higher than Europe for all these top AI roles just isn’t true. Even the big American companies have been opening offices in places like London to hire the top
104.
by BitsAndObjects 7mo ago
You mean because of the international sanctions that needed Taiwanese, British and Dutch support to be effective? Or because of the revoked processor design licenses from the British company Arm (which is still UK headquartered… despite bei
105.
by BitsAndObjects 7mo ago
I sometimes wonder whether people realise which country ASML is based in, and which country their major suppliers are in (e.g. optics: Germany)
106.
by BitsAndObjects 7mo ago
For sure we’re not currently attracting the talent. There’s more to that than just money, but money is significant factor. When it comes to compensation, AI is too broad a category to have a meaningful debate. Hardware or software or mathem
107.
by BitsAndObjects 7mo ago
You seem to have a very ill-informed view of UK/EU salaries in this particular sector; And also: yeah, people take salary hits to go do things they believe in (this is like, the entire premise of the underpaid American startup founder
108.
by BitsAndObjects 7mo ago
Yes. And the US can’t realistically stop our well-funded homegrown AI Hardware startups from manufacturing with TSMC. This is part of why there’s funding from the EU to develop Sovereign AI capabilities, currently focused on designing our o
109.
by BitsAndObjects 7mo ago
The UK and Europe welcome the US Footgun Operation. Plenty of opportunities for those top researchers and engineers over here. The EU (which is not the same as Europe), is also looking a bit sharper on AI regulation at the moment (for now…
110.
by BitsAndObjects 7mo ago
Sarcastically, it's "the vibes intensifying". (Vibes ~ Vibrations ~ Heat) Tbf it's a reasonable question... I think it's a little tricky to pin down the equivalent of "kinetic energy" in purely economic te
111.
by BitsAndObjects 7mo ago
Circular-breathing causes the air to heat up, causing expansion. This is how a balloon can expand even when someone is breathing air from inside it. s/breathing/investment/g s/balloon/bubble/g s/air/m
112.
by BitsAndObjects 7mo ago
You may also be interested in: https://go.dev/blog/greenteagc
113.
by BitsAndObjects 7mo ago
See also: https://go.dev/blog/greenteagc
114.
by BitsAndObjects 7mo ago
I'm not saying I do this but if I were as smart as I think I am I would have given a Gmail example rather than the example you've given to avoid bots just looking up my website and starting to bypass my setup... ;) ;) ;) Also, spa
115.
by BitsAndObjects 7mo ago
You'd have thought so, but no, in my experience this works very well. People doing this kind of spamming don't seem to be particularly bright, nor do they seem to spend any time/effort to clean up their scraped database.
116.
by BitsAndObjects 7mo ago
N.B. Using service-specific emails is trivial - you don't need separate email accounts. Just use email aliases, e.g. "john.smith+github@gmail.com" -- which is an alias called "github" for "john.smith@gmail.com&
117.
by BitsAndObjects 7mo ago
My solution to this is to use a Github-specific email address. All emails sent to that address which do not originate from GitHub are immediately reported as spam, marked read and deleted. I sometimes use different git/GitHub addresses
118.
by BitsAndObjects 7mo ago
That would be a fundamental change to how Git works, not just GitHub. Even if the web UI didn't show it, a simple `git log` would reveal it. You can mask your email address in git commits but a lot of open source projects won't ac
119.
by BitsAndObjects 8mo ago
The oil industry? (Only half joking)
120.
by BitsAndObjects 8mo ago
The article also ended by suggesting trial work periods. I believe software engineering basically has these in the form of "internships". At least, at a lot of UK companies, internships lead to jobs (often after graduation). Not t

More ›