smokefoot

5 ms·

1.
by smokefoot 21d ago
This has probably been posted before. As interest in hardware acceleration for AI intensifies, I started to think about whether analog circuits would make more sense. You can imagine actually modeling neural network architecture in the circ
2.
An analog-AI chip for energy-efficient speech recognition and transcription(nature.com)
4 points by smokefoot 21d ago|1 comments
3.
by smokefoot 21d ago
Yes. Each industry has developed accounting standards that reflect the nature of their business. You couldn't run a bank or a payments company with a simple sales - COGS = gross profit model, it just wouldn't make sense. In health
4.
by smokefoot 3mo ago
Yeah but the smoking gun is that volumes didn’t really go down much. I would buy higher unit margins on a shrinking volume, but for total profits to materially increase that means prices went up without a significant impact to volume. That
5.
by smokefoot 3mo ago
Woodworking taught me a lot about planning and design. As a young person, I was like the authors brother. I just wanted to do the thing, not draw a diagram and figure out how much wood I need, or build a fixture to mark the stair lines. Woo
6.
by smokefoot 3mo ago
Yes but what about AI? (Perhaps the most annoying words written in the last few years mostly on LinkedIn). But actually in the years since this was written, I do think the world has shifted. Doing things on a computer used to be really hard
7.
by smokefoot 3mo ago
Doubt it. Consulting firms run 30% gross margin and operating margins in The single digits or teens. I’d bet IBMs legacy portfolio is 40%+ operating margins. Sometimes you might hear private partnerships quote gross margins in the 40s or 50
8.
by smokefoot 3mo ago
I mean no. The LIBOR analogy is appropriate. Large, long-term egg supply contracts are fixed to an index and that index was manipulated. That's criminal conspiracy and price fixing, not just a liquid market. That's notably differe
9.
by smokefoot 3mo ago
Maybe fair. I think my point was the author emphasizes how strange the software is. The further you are from the training data, the less well a model will perform. I haven't looked at the project, but it seems like it could maybe be wr
10.
by smokefoot 3mo ago
The author admits that the logic of the language and the design of the parser are idiosyncratic. Even the solution the author likes is an extension of an existing hacky trap door. He could be more open-minded about the solutions the AI prop
11.
by smokefoot 3mo ago
Which is so weird, right? Like what is IBM now and how does a research lab make sense with the rest of their business? The money-making parts of IBM are: legacy software and hardware (declining), consulting (low margin, low leverage), enter
12.
by smokefoot 3mo ago
Um, did i just get mansplained compound interest by Paul f*ing Graham? I feel like this has been the subject of condescending advice since the beginning of time. "But now you at least understand, from having done the math yourselves, t
13.
by smokefoot 1y ago
Or intel … but I think Samsung comes considerably closer. I’m not close to it.
14.
by smokefoot 1y ago
I mean, I don’t know how long the NVIDIA moats can hold. With this much money at stake, others will challenge their dominance especially in a market as diverse and fragmented as advanced semiconductors. That’s not to say I’m brave enough to
15.
by smokefoot 1y ago
Chinese semiconductor dominance is not imminent and US containment has been somewhat effective. I don’t think that will hold on a generational timeline, but it will be hard to overcome.
16.
by smokefoot 2y ago
It’s pretty hard to quantify. Reading the tone of NYT headlines over the last 20 years, I observe a steady increase in side-taking on both sides of the newsroom. Those headlines and stories are empirically more engaging and the market value
17.
by smokefoot 2y ago
Yeah, I mean I’m being negative but I enjoyed the video. I just found Ben’s whole line of questioning to be pandering and not a dialog (“why not MIT?”) I had a similar experience with BBSs and I remember the romance of using a modem. I actu
18.
by smokefoot 2y ago
What a circle jerk. I guess there aren’t successful people with any humility. But seriously, he just used his own podcast to feature himself!
19.
by smokefoot 2y ago
The industry moved to 300mm 25 years ago. It’s going to take a lot to get off that standard. Also, round chambers for etch and deposition are good for homogeneity. I can imagine square chambers would result in lots of process challenges.
20.
by smokefoot 2y ago
The cost of silicon is less than insignificant in an advanced semiconductor product.
21.
by smokefoot 2y ago
It’s interesting to see the give and take between google sheets and excel. Google sheets came on the scene shooting for total backwards compatibility and then proceeded to develop some really interesting innovations. Now we see new features
22.
by smokefoot 3y ago
Right, good point. We will see if regulators take up that cause.
23.
by smokefoot 3y ago
The secrecy is ick, but this is the future and there’s no stopping it. There’s ample evidence that consumers won’t pay for privacy and as most consumers opt in to data sharing programs, the non-data-sharing cohort will get seriously adverse
24.
by smokefoot 3y ago
This is not how capitalism works. However, if insurance were priced perfectly, it would cease to be useful!
25.
by smokefoot 3y ago
There used to be opt-in insurance programs with many carriers. They used to send you a device, but I guess that was mooted by secret mass surveillance?
26.
by smokefoot 3y ago
Some serious ADHD apparent in this interview. Like what was that digression about North African countries writing in French? That really should have been edited out.
27.
by smokefoot 3y ago
Most carriers operate both Airbus and Boeing. Low cost carriers have less diversity and operate either 737s or A300s. Southwest is the only airline I can think of that would have their hand forced in this way.
28.
by smokefoot 3y ago
Also there’s less downside risk for OpenAI. Google has layers of approvals and risk committees because they don’t want to put the money machine at risk through litigation, reputation or regulation. OpenAI has nothing to lose—this is their o
29.
by smokefoot 3y ago
Yes! And aggressively enforce your software license agreements with frequent audits. Anything to make money other than innovating and actually building better products.
30.
by smokefoot 3y ago
It’s not that different from software focused private equity shops like Thoma Bravo. Buy a software company, starve it of resources, and price just below each clients BATNA. Revenue may decline over time, but costs fall faster and the buyer

More ›