dofm

22 ms·

511.
by dofm 2mo ago
Indeed. I like using Macs mostly, and the bargain M1 Max MBP I am using for local LLMs is a fabulous experimentation platform and does loads of other stuff well, so I am in no rush, but if I reached the point of buying dedicated hardware fo
512.
by dofm 2mo ago
This really very much depends on how you are using it, I think. If you intend to leave it to solve long context problems and write whole prototypes, the 27B is going to be much better. But if you are sort of pair-programming with the model,
513.
by dofm 2mo ago
I think the big picture is whether the big two can mask the real cost of tokens (inference and training) long enough to survive; that OpenAI is seeking government investment suggests that they can't. Ed Zitron helps you see the broad a
514.
by dofm 2mo ago
Rails was fully into growing pains and maintainability crises (some large rails codebases took years to migrate) and PHP was in transition; some good things by then but it was not what it is now.
515.
by dofm 2mo ago
You really do have to be well under thirty years of age to think 2035 is far enough away for any of this stuff to be plausible. Ask anyone over fifty how far away 2035 is and they will tell you something like it's not long after Christ
516.
by dofm 2mo ago
I think it was fair. I added the bit in brackets to clarify because maybe it had come across as a nit-pick. In my head there are corrections that are rude that I don't do, because I genuinely think authentic writing matters more and mo
517.
by dofm 2mo ago
“TK” meaning “to come” is another one. “More TK” in particular is now a blogger affectation but it seems to have been a way to communicate to an editor that a story was so far incomplete, within the text of a submission, without the editor
518.
by dofm 2mo ago
Random trivia (and something I learned this way myself from an internet reply, though I can't remember when): What gets buried is the "lede". It is a really unusual thing — a deliberate, engineered misspelling of "lead&q
519.
by dofm 2mo ago
We are saying the same thing except for that specific narrow point (I edited my post a little to make it clearer). I was just observing that if that is a consensus on the continent — that we're going to be begging to be let back in on
520.
by dofm 2mo ago
Nobody pro-EU left in the UK believes we will be able to rejoin with the rebate, veto or retaining our currency. I don't know who is telling you that we think this: you should probably not listen to them because they are relitigating o
521.
by dofm 2mo ago
It happened because we left border treaties that meant that continental EU states were obliged to prevent onward travel to us. Once we stepped out of that, we needed to find new incentives for France and Belgium to care. We have failed to d
522.
by dofm 2mo ago
Indeed.
523.
by dofm 2mo ago
Indeed. Even in that speech he contextualises it as a long-term goal. Plus, what he thinks and what is possible are different things, and he absolutely knows it. We will, I believe, not rejoin the EU while it is in its current form, because
524.
by dofm 2mo ago
Is this really true? At least one very headline AI researcher is pretty much an Anthropic spokesperson.
525.
by dofm 2mo ago
I also find that the 35B A3B in Q4 outdoes Ternary Bonsai 27B. Mind you, because I am not that interested in long-context, fire-and-forget tasks, I think the faster MoE model has more value generally.
526.
by dofm 2mo ago
The real target of small models is surely the opposite: knowing with clarity how to conversationally or otherwise reason about your requirements in natural language, sound and imagery, consult APIs that interact with apps, and then have a s
527.
by dofm 3mo ago
He doesn't pretend it's not going anywhere. He does think the market is going to crash. Like I say, I find his tone enjoyable, some of his predictions are interesting (and he has already been proved right on its risks to Oracle
528.
by dofm 3mo ago
> His value as a prophet of the AI bubble is going down by the day Is it? I don't think this is really true. For example he's spent some time talking about Oracle's relatively more dangerous exposure to OpenAI. Here he is
529.
by dofm 3mo ago
> The bullish AI side is full of grifters and folks that were block-chain and NFT “experts” before they became AI “experts.” At the beginning of this whole thing, when I still had an X account to log in with, I used to scroll back on an
530.
by dofm 3mo ago
If it gets a distributor!
531.
by dofm 3mo ago
> You clearly don't like the idea there are some businesses in the AI mix that will continue making money and/or diversify their divisions to cover for any losses they suffer during this hypothetical crash or I wouldn't be
532.
by dofm 3mo ago
Fully possible to square those two things. Like, if you were young and you knew even trivial things about how much your parents or relatives earned, you might have thought “how can they afford this new house” but then see others getting new
533.
by dofm 3mo ago
> So in a couple years now, AI will be as gone as cabbage patch kids and pet rocks? > IMO that's anti-AI Psychosis, the evil twin of believing GPT-4o was sentient. IMO that’s a particularly scraggly straw man. Even Ed Zitron, who
534.
by dofm 3mo ago
Ed Zitron is enjoyably mouthy and rude about AI and AI people, but it’s a bit of a stretch to suggest that someone most people have never heard of helped create the AI backlash. He is just putting it into words. You know what created the AI
535.
by dofm 3mo ago
Interesting though that the 1.5 setting (at least on my older iPhone) doesn’t really ever seem to cover the range of skin tones of indigenous Australians (“aboriginal” and Torres Strait Islanders), who apparently get their skin tones from a
536.
by dofm 3mo ago
Serious question: are you having a particularly bad day or week? Because this was really rude and you appear to have poured out a lot more irritation than was merited. If it were not for this, as a geeky photographer I'd get into it wi
537.
by dofm 3mo ago
It is all of those things and more. Ethical: favour the notion that humans should be who we look to for writing, and that some of them are hopefully meaningfully compensated for it Political: favour the decentralisation of effort away from
538.
by dofm 3mo ago
There is no "before that", not on the internet. Stock photography as a business is more than a century old, and we were using it on the internet as soon as you could put an IMG tag in a page. Lazy use of stock imagery is a probl
539.
by dofm 3mo ago
I hope so — I find the 3.6 35B to be well worth using.
540.
by dofm 3mo ago
> I can’t remember of DK suggests some sort of effect where the expert has undue self-doubt, though… There is a brief bit where they attribute that observation to other causes: In making these predictions, we felt that we could account

More ›