dekhn

24 ms·

571.
by dekhn 1y ago
also known as the 'double down on stupid' and 'triple down on stupid'
572.
by dekhn 1y ago
Issues with master locks are hardly new- back in the 1980s, I downloaded a file from a BBS explaining how to open a combo lock (basically by pulling on the shackle while turning, and a few other tricks. It's still online: https:/
573.
by dekhn 1y ago
Read more about the author: https://en.wikipedia.org/wiki/Charles_Murray_(political_scie... I do not find anecdotal evidence he provides as being a convincing case for a non-materialist soul.
574.
by dekhn 1y ago
I helped Google's genomics/health projects get started many years ago and literally the top item of my docs was "do not claim you are going to cure cancer, because the medical establishment knows that's not an achievable
575.
by dekhn 1y ago
A lot of us grew up in the time of 3 Mile Island and Chernobyl. Combined with a lot of media about radioactivity's dangers ( https://www.nytimes.com/1999/06/06/nyregion/babies-teeth-and... ). I think
576.
by dekhn 1y ago
I was talking about things inspired by (for example) hidden markov models. See https://en.wikipedia.org/wiki/Graphical_model In biology, PGMs were one of the first successful forms of "machine learning"- give
577.
by dekhn 1y ago
What exactly are you suggesting- that the SAT solver example given in the paper (applied the HP model of protein structure), or an improvement on it, could produce protein structure prediction at the level of AlphaFold? This seems extremely
578.
by dekhn 1y ago
We have an open office surrounded by "breakout offices". I simply squat in one of the offices (I take most meetings over video chat), as do most of the other principals. I don't think I could do my job in an office if I cou
579.
by dekhn 1y ago
Many researchers who invent something new and powerful pivot quickly to something new. that's because they're researchers, and incentive is to develop new things that subsume the old things. Other researchers will continue to wo
580.
by dekhn 1y ago
Genomics, protein structure prediction, various forms of small molecule and large molecule drug discovery.
581.
by dekhn 1y ago
The way I look at transformers is: they have been one of the most fertile inventions in recent history. Originally released in 2017, in the subsequent 8 years they completely transformed (heh) multiple fields, and at least partially led to
582.
by dekhn 1y ago
Thompson and Pike: https://en.wikipedia.org/wiki/UTF-8 """Thompson's design was outlined on September 2, 1992, on a placemat in a New Jersey diner with Rob Pike. In the following days, Pike and Thom
583.
by dekhn 1y ago
This is the canonical example: https://www.nytimes.com/2023/05/22/technology/ai-photo-label... Some more details in https://www.nytimes.com/2021/03/15/technology/artif
584.
by dekhn 1y ago
It's entirely unclear to me if a system the size and scope of AWS could be re-thought using these principles and successfully execute a complete restructuring of all their processes to reduce their failure rate a bit. It's a syst
585.
by dekhn 1y ago
When I was an SRE at Google our oncall was extremely serious (if the service went down, Google was unable to show ads, record ad impressions, or do any billing for ads). It was done on a rotation, lasted 1 week (IIRC it was 9AM-9PM, we had
586.
by dekhn 1y ago
I worked with a guy who advanced the state of the art for this- https://en.wikipedia.org/wiki/Crooks_fluctuation_theorem which was used to measure breaking individual bonds while unfolding RNA: https://pubme
587.
by dekhn 1y ago
I only started managing people recently (and still do some engineering and development, along with various project management- my job title is "Senior Principal Machine Learning Engineer - so not really even a management track). I have
588.
by dekhn 1y ago
Google had been burned badly in multiple previous launches of ML-based products and their leadership was extremely cautious about moving too quickly. It was convenient for Google that OpenAI acted as a first mover so that Google could ente
589.
by dekhn 1y ago
but X isn't really an insulated org... it has close ties with other parts of Google. It shares the corporate infra and it's not hard to get inside and poke around. it has to be, because it's intended to create new products
590.
by dekhn 1y ago
Well, let's give a concrete example. I want to use an SaaS as part of my job. My manager knows this and supports it. In the process of me trying to sign up for the SaaS, I have to contact various groups in the company- the cost cent
591.
by dekhn 1y ago
[2018]
592.
by dekhn 1y ago
Sometimes moving fast in a large org boils down to finding a succinct way to tell the lawyer "I understand what you're saying, but that's not consistent with my understanding of the legality of the issue, so I will proceed wi
593.
by dekhn 1y ago
Yes, I was SRE at Google (Ads) for several years and that influences my work today. SRE was the first time I was on an ops team that actually was completely empowered to push back against intrusive external changes.
594.
by dekhn 1y ago
it's more "move fast on a good foundation, rarely breaking things, and having a good team that can fix problems when they inevitably arise".
595.
by dekhn 1y ago
I'm seeing a lot of frustration at the leadership level about product velocity- and much of the frustration is pointed at internal gatekeepers who mainly seem to say no to product releases. My leadership is currently promoting "be
596.
by dekhn 1y ago
I recommend reading "Operation Paperclip" to see the conditions the Nazis put their concentration camp workers in. Not really isomorphic.
597.
by dekhn 1y ago
I always laughed at that line from Lehrer, but Von Braun absolutely cared about where his rockets came down.
598.
by dekhn 1y ago
National pride has long been tightly coupled to seafaring capabilities. https://en.wikipedia.org/wiki/Vasa_(ship) "Richly decorated as a symbol of the king's ambitions for Sweden and himself, upon completion
599.
by dekhn 1y ago
A diffuser is basically applying a convolution to the underlying pixel array (moving the diffuser away from the LED increases the convolution width, but unfortunately also reduces the intensity). I don't think you can exploit that to s
600.
by dekhn 1y ago
Are you using USB-C for networking between the Spark and the Mac?

More ›