cheald

Abliteration and fine-tuning remain standard operations on open weight models.

For what it's worth, the Navy method (https://www.omnicalculator.com/health/navy-body-fat) always matched my DEXA scan results much more closely than any BIA scale I used ever did. Given that, it would make sense that you could estimate it from photos.

A robust open source profile is my single favorite hiring profile indicator. However, with the current state of things, if I get a whiff of AI-driven "contribution" it becomes an instant black mark against the candidate.

So far I've been impressed enough with the HW4 Teslas that I haven't had them do anything that I had to intervene to correct or prevent. It's pretty amazing at how well it handles all kinds of things - construction, weird merges, road debris. This morning, there was a tire in the middle of the road, which caused traffic ahead of me to slam to a halt. Mine had to brake hard enough that ABS engaged, and then navigated around the tire. I was impressed.

Very fun! There seems to be an exception when using the 10-roll function that might need some attention.

    index-BwvgcC1c.js:62 Uncaught TypeError:
      C is not a function at zi (index-BwvgcC1c.js:62:150798)

Those lines of code are:

    for (var oe = [], ce = 0; ce < 10; ce++)
      oe.push(C());

"Applied theology with inadequate safety margins" might just be my favorite difficulty description ever.

If you wouldn't give it to an enthusiastic junior dev, don't give it to AI, period.

Hang out at a BJJ or MMA gym for a bit, and you'll find plenty. Peptides are really popular in combat sports circles, with good reason.

I used a combo of low-dose retatrutide, tesamorelin, and ipamorelin and lost about 15lb over 45 days, including 60% of my visceral fat, and put on 4lb of muscle, per before-and-after DEXA scans. I lifted regularly, ate well, and prioritized protein, and while I definitely under-ate protein, I was very pleased to find that I was able to increase muscle mass while cutting the fat. My visceral fat was the primary target here, since I'd been unable to get it to budge despite consistent training and diet. Very pleased.

It's actually very easy to cool in deserts, because low humidity makes it very easy to move heat into the ambient air. You have to contend against ambient temperatures, but that's what insulation is for. The other big things you need for datacenters are reliable power and a low probability of infrastructure-disrupting natural disasters.

I've done a lot of exploratory work with Stable Diffusion LoRAs, and I actually do buy that there's some juice here, though it's almost certainly not nearly as good as other techniques can be. In particular, this technique will likely avoid the intruder dimension problem which plagues naive LoRA. SVD is expensive, but you only have to do it once at the beginning of training.

I haven't done much research lately, but when I was working on it, I was having substantial success training an adapter of the form U_k @ P @ A, where U_k was the top k left singular vectors of the underlying weight, and then P and A were your typical LoRA projection matrices.

The 13 parameters are kind of misleading here; the real juice is going to be in the P_i fixed random matrices. My suspicion is that they are overfitting to the benchmark, but they almost certainly are observing a real gain in model capacity that is largely due to avoiding the intruder dimension problem.

Have you not seen Instagram lately? It's overwhelmingly clogged by AI content, much of it masquerading as real people.

It's best to think of instruct-tuned LLMs as mirrors rather than intelligences. They generally reflect what you're putting into them, but they do it in a way that can easily masquerade as wisdom. I think this makes it really easy for people to self-delude.

Remember that many people are heavily are happy-path biased. They see a good result once and say "that's it, ship it!"

I'm sure they QA'd it, but QA was probably "does this give me good results" (almost certainly 'yes' with an LLM), not "does this consistently not give me bad results".

I guess mine hasn't updated yet. I'm a paid user, but I'll be migrating elsewhere. What a disappointment.

Nova has been my favorite launcher for years, but after this, I may have to look elsewhere. Even as a paid user, I don't have much confidence that I'm not being sold off for ad exploitation.

Technical analysis is the projection of future price data through analysis of past price data (usually for the purpose of trying to create trendlines or find "patterns"). Options pricing is quite a different beast - it encodes marketwide uncertainty about the future price of the underlying, which has little to do with the past price action of the underlying, and everything to do with all known information about the actual underlying company, including fundamentals analysis, market sentiment, future expectations and risks, etc.

To put it another way, to price an option I need a) the current price of the underlying, b) the time until option expiry, c) the strike price of the option, and d) the collective expectation of how much the underlying's price will vary over the period between now and expiry. This last piece is "volatility", and is the only piece that can't be empirically measured; instead, through price discovery on a sufficiently liquid contract, we can reparameterize the formula to empirically derive the volatility expectation which satisfies that current price (or "implied volatility"). Due to the efficient market hypothesis, we can generally treat this as a best-effort proxy for all public information about the underlying. None of this calculation requires any measurement or analysis of the underlying's past price action, patterns, etc. The options price will necessarily include TA traders' sentiments about the underlying based on their TA (or whatever else), just as it will include fundamentals traders' sentiments (and, if you're quick and savvy enough, insiders' advance knowledge!) The price fundamentally reflects market sentiment about the future, not some projection of trends from the past.

This isn't technical analysis, this is an article on how to use the options market's price discovery mechanism to understand what the discovered price implies about the collective belief about the future price of the underlying.

The entire options market is built on this kind of analysis.

Thank you so much for this. I am absolutely gonna pick up a couple of these for my kids. What a great way to offer creativity on demand!

Arc Raiders runs great. ProtonDB is the right thing to check to find out if any given game is gonna run on Linux. Fortunately, the success of the Steam Deck has more and more devs playing ball.

You could just buy another SSD and install Linux on that. Then, you have your Windows drive left untouched and pristine so you can swap back if you want, or you can pull data over as needed.

"Local man uses AI to try slightly different casserole recipe" just doesn't have that click-driving wow factor.

Yogurt's been a staple of mine ever since I started eating clean years ago. It's high in protein, can be mixed with all sorts of protein powder, fruit, nuts, and seeds, and is always quick and easy to prepare. The probiotic effects are an excellent benefit, but the real trick for me is that it offers that junkfood-like "quick snack" availability while being actually pretty nutrituious. It makes a pretty solid default snack.

You ever shop hungry? You buy way more, and often not the stuff you should be buying.

If the primary shopper in the household is on GLP-1, I would expect the household expenditures to drop.

I appreciated the essay, GPTisms aside. The core concept is one that I've felt for a long while, but you articulated it more cleanly than I've been able to. I'd explained it as into "the tasks are replaced, the job is not", but I like your distinction better.

I've rather suspected that people are subconsciously adapting their language patterns to those that they hear over and over, and with AI content so prolific online now, it's natural that people are being programmed more and more with those patterns.

We trained a model on human language which is now in the business of itself retaining human language.

Oh, I like this a lot. My kids are quite physically active, but they do love to binge video games, too. I like the idea of letting them "buy" more leisure time at their own discretion through self-disciplined work.

It's not entirely devoid of value, but that doesn't make it a good idea. Junk food contains healthy ingredients, too.

I receive several "Let's help get you back onto Instagram" emails a week, and have for months and months. I can only assume it's someone trying to do something nasty, but I have no idea what it actually could be.

It's quite perplexed me.