deepGem

> Agreed. I was excited by the prospect of being a part of a network of other vetted founders.

I feel compelled to say this. I am one of the founders who got an initial reject followed by an accept email. I read the accept email and felt good (I had not even read the reject email). When I learned that YC has accepted everyone, I felt better. Sure having access to a dedicated advisor has an advantage. The advisor can guide you and all that but it's ultimately your hustle. Now the playing ground is more or less levelled when it comes to hustle.

Also, how does this 'part of a vetted group' feeling help in your startup's success ? Agreed, it's a momentary boost in confidence, but beyond that I'm not sure how the vetting process adds any value. The way I see it, the vetting is more or less random.

"Our team spends hours in the kitchen tweaking each recipe so we can prep and cook it just enough"

This is super tricky to achieve. But, are you sure that this process is what makes your food flavoourful yet preservable by just freezing/refrigerating ?

Good luck though!

I hope you mean the stock.

You can use www.breezometer.com to check real time PM2.5 readings in your location. Based on that you can figure out whether you need an air filter or not.

This can also have an exponential impact on financial fraud and anti money laundering. Shared fraud data can significantly thwwart or hinder ongoing fraud rings and such. Banks do not share fraud data today precisely due to the lack of such an algorithm.

">By having its own silicone, Tesla can build for its own needs at its own pace ... "

This is not what Musk said on the conference call. Surely some bonehead reporter is writing this. Musk said the latter, precisely to go from current 200 fps speed to something like 2000 fps (I might have the numbers wrong, but it's a 10x improvement) and that they have been doing this in stealth for the last 2-3 years. They expect the cost of the new chips to be the same as that of the current nVidia lineup. That got me thinking though. What are they currently using - the Titans ?

For me, it was the mobile interface interlaced with so many video ads that killed my usage. I do open the app sometimes to check what some of my distant friends post, but my usage has drastically gone down. I think 5 years back I spent at least 30 mins on the app every day, now I don't even open it on a few days. To top that, the Cambridge Analytica incident.

That said, I remain invested in FB, since my hypothesis was that 'a majority of people are not like me'. Looks like I am being proven wrong, but let's see the long term impact.

It's the former. They can be made stable at super conducting temperatures but not completely. There will always be errors. From what I can understand you cant fully eliminate the errors. I donno may be some new superconducting temp might make them stable but that's a wild guess.

So the number of raw qubits are in the order of a million for large prime factorization or to make Shor's algorithm a reality. This is because the physical qubits are highly unstable and prone to errors. The error correction methodologies will result in a fault-tolerant logical qubit. Depending on the error correction algorithm, the ratio of logical: data/physical qubit is 1:1000 at the minimum. You need 4000 logical qubits at least for large prime factorization give or take 4 million raw qubits.

Well, for starters we are looking at question answer type natural language processing, something superior than Google smart replies. So imagine a flow something like '<Bot>, schedule a meeting with <p1> at Noon tomorrow. Now the <bot> has to figure out the availability of <p1> at Noon. This is relatively simple if <p1> has given the bot access to the calendar but that's unlikely. So the bot has to start a series of dialogs which to figure out the availability. Now you are looking at NLP + some state maintenance and it gets even more trickier if you have more than one participant, if there are no available overlaps, if the bot has to force a schedule which is dictated by the boss. Using a phone app you can solve some of these by prompting custom keys but then you run into app distribution challenges.

Thank you. The latter point about accounting metrics that you mention is very valid. I will think on these lines. I am very much interested in reaching out to VCs who are supportive and understand training costs for AI. If you know of any please do point out.

Am glad this is highlighted now. Back in 2015, when we wanted to build an meeting scheduling bot, we naively thought we could use only machines to get the job done. 3 months later we realized that was no way feasible, not then not in the next 10 years. So the common feedback we got was to just use low cost labor in India/Phillipines to get the job done. To us, that was a no-go because we kept privacy as the top criterion for doing anything with people's inboxes. Even after obfuscation, we didn't want any human to read or parse soeone's private messages. So we dropped the idea and shut shop. Sure we failed, but at least we were true to our conscience.

Another way to have tackled this problem was to just go down the research route and build a product once the ML was fully baked, but VCs won't fund such adventures.

To get VC funding, we had to lie through our teeth, which was a total no-go. I am not bringing the righteous argument, but for once I feel really good that our viewpoints held then are corroborated now, some kind of confirmation bias kicking in.

“Once two men share a woman, they’re brothers.”

Is this for real ? WTF

I think we have to stop looking at productivity as a universal metric. Each person defines his own way of being productive. Some people subscribing to deep work only consider deep work as productive work. If I can just write code in my comfort zone for 2 hours should I consider that productivity ? I'm not sure.

From my experience, a lot of a founder's time is spent in non deep work. Taking calls, traveling, talking to users, understanding user feedback, writing product features and coding up some or most of the features. So the actual deep work is perhaps very limited - in desiging features and coding them up. How do you measure founder producivity when the tasks are so varied. IMHO such a measurement exercise is futile.

Where is the need to spend 5.2 billion on public transit ? A few electric buses with point to point connections will do. Start small and expand as needed. Just one high density suburb <-> Downtown connection is sufficient to start with. Electric buses are so good, they have instant torque - so there is none of the lag that plague traditional diesel buses, they are fast, quiet and since it's point to point, your range anxiety is practically eliminated. A dedicated bus lane or the HOV lane and a stop every 1/2 ecits should cut it.

Sure it's not as fast as a dedicated subway or the Shinkansen but at a fraction of the cost, you are bringing in people to start using public transit.

Man, I am appalled at your response. What we are seeing in Delih is beyond alarming. In fact, even the so called normal in Delhi is alarming by the global pollution standards. Please avoid making such flippant comments.

PM2.5 is at a hazardous level in Punjabi Bagh.

http://aqicn.org/city/delhi/punjabi-bagh/

I am an Indian and I echo your sentiments. Federalism, which was supposed to work really well, is failing at all levels in our country. A single vision to move forward does not germinate and even if it does in one person's mind, it is impossible for that vision to see light of the day. Even a flagship project like the bullet train is unable to take off because of corruption at the grass root level (panchayaths and the like).

In India, a vast majority of the population doesn't realize the growth by letting others grow concept. The predominant sentiment is to grow at someone else's expense, meaning I have to grow and others shouldn't or even if they did they have to grow less than me. This manifests in every step and walk of life. This leads to all sorts of corruption. Some guy who owns a key piece of land won't sell that land for a public project, just because a politician who doesn't want that project to succeed will back him up and prevent him from selling. This happened in Bangalore when a key arterial road was being developed. As soon as word got out, some local politician flooded a key government land with slum dwellers and overnight they occupied that land and prevented the road construction. Stories like this are the norm in any land acquisition.

Of course there are exceptions, and that's how this country is seeing modest growths of 6-8%. I say modest because, our baseline is fairly low, as compared to other economies. I feel so bad, we have the highest inhabitable land mass of any country, some of the most fertile regions, no adverse weather in most parts, bio-diversity to die for. It's like someone handed this climate on a platter, yet we don't realize that. We are just destroying this paradise day by day.

Coupled with website notifications, it's all hell break loose. Can someone tell me if it is possible to write a chrome plugin that will close any such websites. I mean within 30 seconds of opening if there is any permission/sign up popup - just close that website.

I use customised resumes too and not having a desktop MS office license makes this process even harder.

Thanks a bunch. I will start doing this right away.

I have been following learning to learn on Coursera. The course instructors lay a strong emphasis on deep work, but teach you how to learn. Prior to taking this course, I was getting into periods of deep work on and off but didn’t learn anything substantial. What I mean by learn is that I couldn’t transfer the learned skills easily.

Now, I spend a lot more time learning a topic, and follow the course methodology of chunking. 2 days in and I can already notice the differences. I woke up this morning realising there is a bug in my queue implementation.

Middle aged people tend to make better polar explorers

Perhaps for the same reasons they tend to be better entrepreneurs.

Almost all estimates are based on past experience. So good luck with such estimates when you are entering new domain or even new tech. I honestly believe that there is no need for estimates. Your delivery date is set by the customer or the market or competition. So it's a lot healthier to work backwards from that date and figure out which of the features can be delivered, than to breakdown the work into quite meaningless chunks, figure out the estimates for those chunks and god forbid, build a Gantt chart.

As someone who worked very closely with tech support and sometimes fielded calls, I echo this. It's batshit insane to field support calls, L1 especially. Honestly, no human deserves the wrath of support calls. You get yelled at for no apparent reason and all you can do is be patient. Who knows what the psychological consequences are. I have seen customer support people lose their temper in otherwise normal situations.

Awesomely done ! What i really like is that, in addition to the ease of deploying a model, this product also lets you visualize the activations of the neural network. I mean you could build the visualizations in Jupyter but the ease of toggling the layers, such as switching of Max-Pool etc is super helpful for understanding how neural nets work.

I once wrote a packaging script that saved 2 man weeks of recurring work. Not the swankiest or the most challenging job I did, but it was immensely gratifying, since this time was directly billed to customers. So yes, fixing small non-critical issues can be very rewarding.

This is a problem even in tropical countries such as India and Singapore. The contrast between outside and inside temperatures is insane. We get headaches and God knows what the long term detrimental health effects are. It's worse in Singapore as the subway is also super cold.

As far as I know this isn't a problem in India, UK and US. Reporting additional income for tax purposes is fairly simple and there is nothing legally wrong in doing multiple jobs at the same time, unless you are on some kind of Visa which prohibits that kind of employment.

Agreed. I guess you could make the contract route optional. I mean have the standard interview route or the contract route - ask the candidates to pick what they want.

Exactly. Just a week or two of contract work will tell you if the candidate is a fit or not. A 30 minute phone screen will tell you if a candidate is worthy of a contract. I think companies don't do this because a. They don't want to pay and b. They don't get to gloat 'we have a super rigorous hiring process'