got2surf

Find an un-sexy domain that you have more access to than the average person. Start to build domain expertise in that area as quickly as you can (people are surprisingly willing to talk when you don't want to sell them something, but just learn about how they do things).

Build something (anything!) that uses some cool ML technology (or even just some simple statistics) in that domain. Think of this as an ultra-MVP - something you build part-time in the order of weeks, rather than full time over months. This keeps your "investment" low and makes you mentally ready to accept changes.

Loop back with the people in the un-sexy industry to get feedback. Remember, not all industries are bombarded with technology - you'll need to strike a balance between showing them something sufficiently "fancy" to pique interest, and abstracting away your technology so they focus on a problem it solves.

When my cofounder and I started ~4 years ago, we chose an un-sexy industry that my cofounder knew from years of consulting. This was awesome, but even with his amazing experience, our initial idea for a product didn't take off (which was fine - it doesn't take too many wins to make the weeks of development/sales profitable).

Over the next 4 years, we built another couple of offerings, which sold better than the first. They took hold in various ways, but the insights behind each product were usually driven (or at least influenced) by the people we talked to about our initial product. Which brings me back to the key - build things in a low-cost way and use that to identify tangentially related problems until you think you've found a big enough pain point.

True, but it would still take more time/effort/knowledge to spin up your servers on commodity boards.

Plus, the level of resources for configuring EC2-like servers, and Heroku-esque preconfigured servers reduce the barrier to entry pretty significantly.

Anyone have an idea of how "technical difficulties" got introduced into the process? What comes to mind is (maybe) using a default seed? Re-initializing the number-picking-machine? Am really curious to see what kind of logical bug could lead to duplicate numbers in pairs like September 28 and September 30, and ... September 29 and October 3.

The Honda Element had some great features - it was a perfect "surf and explore" car.

Only thing that kept me from getting one was the fact that the front doors have to be opened for the rear doors to open. Kind of a pain, especially when you're picking someone up in the rain (which happens surprisingly often in Florida!)

The rise in popularity (and price) of Westafalias recently has been interesting to watch! Huge demand (even on $100k pre-upgraded versions), which seems to be driven by younger folks.

"Better" is tough to say - coding bootcamps are focused on getting a coding job, but university CS degrees are not (CS graduates going into other fields [like finance], research/grad school, etc).

If the goal is to get a high-paying coding job as quickly as possible, bootcamps look good. Would be interesting to look at longer-term trends in a few years - career advancement, job type, management level, industry within tech, etc.

I bought mine through a mechanic/dealer, and had them fix the IMS bearing, rear main seals, chain tensioners, and switch to a metal water pump. Those were the main recommended preventative fixes I heard about.

I've seen most prices for replacement significantly lower than $10k - usually closer to $2-3k (https://www.6speedonline.com/forums/996/241595-ims-bearing-m...). Plus there are a few other things that are easy to do at the same time (like rear main seals).

Completely agree. Plus, after initial depreciation, 911s hold their value pretty well (particularly the 1999-2004 996 models). I view a used Porsche (as compared to new) as "90% of the fun for 50% the cost".

I found the non-AMP version easier to read: https://www.wired.com/story/god-is-a-bot-and-anthony-levando...

1/10 false positives doesn't sound too bad to me. Having seen older family members impacted by the disease with little to no warning, a 90% chance at early detection would be incredible. Even with a false positive, the changes you would make (lifestyle changes, brain stimulation, continuous monitoring) would only be beneficial, IMO.

Take-home salary after taxes could be a lot less than $6,750 per month. Health insurance is probably reasonably priced given that she seems to have a government job.

But I do see your point - she could (and many do) live closer, pay more, and sleep more.

Good point - finding domain expertise for product/market fit is super important.

However, that's not something that is solved purely by capital. It also takes a lot of time (in practice, you need potential customers to describe their problem multiple times over multiple meetings, since what they think they want, what you think they want, and what really solves the problem take time to converge). For our company (applying ML/NLP in the customer feedback domain), finding domain experts willing to contribute their guidance took a while, but has been one of the most important things we've done.

I'm assuming they meant deleting the <div class="video-player"> or <video> HTML element itself.

I agree, I don't think this really matters to the individual startup (in that there are probably much stronger factors for success than # of founders).

But it may matter for investors looking at things with the opposite perspective - for example, is it a good/bad heuristic to ignore companies with 5+ founders?

Instead of looking at "% of successful exits that had n founders", it seems more relevant to look at "% of startups with n founders that had a successful exit".

Without knowing the distribution of startups with 1, 2, 3, 4, 5+ etc founders, it's hard to tell how much more/less likely each group is to succeed.

Good point on the terminology. My main (and relatively early) experience has been that there are good lessons to be learned from both sides (like sustainable growth and profit models on the "non-startup" end, and rapid innovation/marketing on the "startup" end)

It begs the question of why we (generally) feel like tech startups need to think/behave differently than traditional businesses.

"Startups" in the sense of new businesses have been around for centuries (arguably longer). But recently, the popular image of most tech companies is raising large amounts of money to grow/scale really quickly. If anything, shouldn't the fact that engineers understand how to leverage technology lead to higher profitability and lower need for capital than all the brick-and-mortar businesses that came before us? Like the SparkFun founder says, as an engineer, I started a company that I knew how to build and implement on my own.

Parent commenter here - I don't doubt that LSD has real effects. I think it's plausible that microdosing has real effects. But I also think that said effects are difficult to measure, because studies typically use subjective mood assessments, which are going to be pretty heavily impacted by the placebo effect.

But even if most of microdosing is just a placebo effect, I think that's still amazing in itself. Even if you're not "physically tripping", you've willed yourself into a "more open" mental state, creating your own positive side effects.

While this is a really interesting article, there may be some issues around the "placebo effect" of taking a microdose vs physical effects of the microdose. (Granted, the impact of "having more creative moments" may be the same either way, which could be a win in itself)

https://www.gwern.net/LSD%20microdosing this article was posted on HN a few years back, and talks about an self-administered blind study of microdosing, which found limited benefits of microdosing, counter to most anecdotal examples.

This is exciting! If you look at historic speeches (ie from American Rhetoric http://www.americanrhetoric.com/top100speechesall.html), there are large variations in average characteristics between various styles/contexts (on average, pitch/volume/speed are different for inspirational vs somber speeches, for example). But there are also really large differences in the variation - an inspirational speech may be marked by large swings from quiet, reflective pieces to booming, rousing calls-to-action while a somber speech has fewer swings in delivery.

For the examples given for various intonations from Obama/Trump, some intonations are much more natural than others. It would be interesting to decide how to parametrize a sentence for the intended intonation. (based on word2vec analysis of the words in the sentence, punctuation cues in the sentence, and perhaps a specified category of "emotional delivery").

It would be interesting at the sentence-level, but also at the macro speech-level to include the right "mix" of intonations for a specific context. On a related note, it would be interesting to study the patterns of intonations in successful vs unsuccessful outbound sales calls, for example, to learn how to best simulate a good human sales voice.

Absolutely - listening through the multiple samples with different intonation from both Obama and Trump, some of the samples are much more realistic, while others come off as robotic.

Maybe it would be possible to train the system to prefer certain intonations in certain cases by rating the realism of the speech in context. It would be interesting to analyzes pauses around words grouped by word2vec! Or choosing a "style" of intonations based on punctuation, parameters like words/minute, etc.

Great point, I think the tech crowd may overestimate the cost of glitches, relative to everything else at play in a business.

I think the point I'm getting hung up on is that the bank's stock price could drop for two reasons: bad PR due to the glitch, and/or falling financials due to fraud perpetrated as part of the glitch. I can completely understand a hedge fund trading and making money off the bad PR. But if (hypothetically) the bank lost a ton of money by hackers liquidating user accounts or, worse, making leveraged bets [before everyone checked for that sort of thing ;)], and the hedge fund knew there was a reasonable chance that the malicious activity would occur based on the newly disclosed information, would they have liability there? (from the theft/fraud perpetrated against the bank, not the drop in stock price)

Great link, thanks for sharing. The quote that stood out to me was “My issue was that patient safety wasn’t front and center.”

I don't have a problem with MedSec making money by shorting St. Jude's stock (that seems to align incentives to take care of security issues as early as possible). But if MedSec publicly disclosed specific, exploitable vulnerabilities (I'm not sure about specifics from the article), they shouldn't be able to hide behind the "doing what is best for the consumer" argument. It's definitely a clever business hack, and that's alright, but the fake sense of moral superiority isn't.

I agree that making bets by noticing public information earlier is 100% okay (and in the case of Lumber Liquidators, a better outcome for almost everyone).

But would this case with the bank be different because the vulnerability, unlike formaledehyde, could be actively exploited? Encouraging a stock price to fall because of bad practices seems alright (like the LUmber Liquidators example), but if in the process you become an accessory to smaller-scale fraud against individual account owners, is it still "alright"?

Just curious, what would the legal implications of something like that be? It seems like you're still benefitting from criminal activity that you enable, but what would the specific charge (if any) be? And any examples where people have tried this?

Although I guess it could help align customer and business goals, since no one wants to lose money

Not sure about octopus, but dolphins do get high on pufferfish - http://www.smithsonianmag.com/smart-news/dolphins-seem-to-us...

"Pufferfish produce a potent defensive chemical, which they eject when threatened. In small enough doses, however, the toxin seems to induce "a trance-like state" in dolphins that come into contact with it"

I agree, and that's a fascinating link. I think that "hacker" can be an accurate/valid description, but also a poor marketing choice at the same time.

Interesting point, but why fight an uphill battle? Wouldn't it be easier to avoid the term "hacker" because of the negative connotation? The importance of "hacker culture" (experimentation, clever solutions, etc) goes far beyond the name.

As someone who reads Hacker News, I understand what you mean by "hacking is essentially a creative process," but wouldn't it be easier to describe it as something like creative security research instead of hacking? "Hackers" will still understand what you mean, and maybe it would be less intimidating to others. Curious to hear other thoughts.

True, but in the article, both Partovi and Ruffin cite examples of clients who "claim they own a house, or that a relative will show up any day now", despite being injured/homeless/exposed.

To me at least, it seems logical that a reasonable person would choose shelter over being outdoors and injured for 15 years. Making that choice is their right, but let's address their mental health problems first to give them the autonomy to make that choice.