m_ke

12 ms·

121.
by m_ke 2y ago
Thanks, I’m actually pretty far in the interview process with a bunch of companies now but would be open to learning more. My contact info is in my bio.
122.
by m_ke 2y ago
Putin would be asking for WW3 if he tried that and it would not end well for him. A ton of big tech companies opened large offices in Poland in the past few years. It’s starting to look like a pretty major tech hub. If you look at Google or
123.
by m_ke 2y ago
Yeah the process these days is insane. I went through 4 rounds of interviews with one company, did really well and was expecting an offer but they want me to come in for a full day of onsite work mixed in with more technical rounds. I’m int
124.
by m_ke 2y ago
It's not just you. I have an Ivy league degree, worked in deep learning since alexnet at a leading startup in the space, was a CTO of a startup that got acquired and have referrals from very senior people at the top FANG companies and
125.
by m_ke 2y ago
It would be great if it wasn't completely wrong 50% of the time.
126.
by m_ke 2y ago
SEO ruined the web, guided by Google's ranking algorithm. Things will get even worse as scammy companies start flooding the web with LLM generated content pushing their products to bias LLMs to increase the probability of outputing the
127.
by m_ke 2y ago
I've seen founder led companies also get derailed by this, usually after raising a large round and getting forced by VCs to put their buddies in management positions.
128.
by m_ke 2y ago
I've never seen a company benefit from layers of MBAs who are only there to hide their screw ups from the leadership while doing whatever they can to get promoted. Strong organizations are usually bottom up, with a lot of ownership and
129.
by m_ke 2y ago
The world is a fuzzy place, most things are not binary. I haven't worked in medical imaging in a while but VLMs make for much better diagnostic tools than task specific classifiers or segmentation models which tend to find hacks in the
130.
by m_ke 2y ago
Yes I am, these LLM/VLMs are much more robust at NLP/CV tasks than any application specific models that we used to train 2-3 years ago. I also wasted a lot of time building complex OCR pipelines that required dewarping / imag
131.
by m_ke 2y ago
It’s the same softmax classifier
132.
by m_ke 2y ago
None of these were binary decisions, but classifying one of around 10-20 conditions or rating cases on a 1-5 scale. In all cases the models trained on a lot of this feedback were more consistent and accurate than individual expert annotator
133.
by m_ke 2y ago
I’ve done a few projects that attempted to distill the knowledge of human experts, mostly in medical imaging domain, and was shocked when for most of them the inter annotator agreement was only around 60%. These were professional radiologis
134.
by m_ke 2y ago
“ OpenAI is a non-profit artificial intelligence research company. Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return. Since our re
135.
by m_ke 2y ago
Don’t make todo lists, use your calendar to plan / time block when you’ll do things and try to stick to it.
136.
O1 tops new aider polyglot leaderboard (61.7% vs. 45.3% for Claude)(aider.chat)
1 points by m_ke 2y ago|0 comments
137.
by m_ke 2y ago
ARC is a silly benchmark, the other results in math and coding are much more impressive. o3 is just o1 scaled up, the main takeaway from this line of work that people should walk away with is that we now have a proven way to RL our way to s
138.
by m_ke 2y ago
This is what happens when you have an economy driven by finance, with corporate overlords who see their 2 million employees as numbers in a spreadsheet that can be moved around to squeeze out as much "shareholder value" as possibl
139.
by m_ke 2y ago
When you have over half a trillion of revenue a year the fractions of a percent add up. Doing this allows them to negotiate a lower price from the manufacturer or get the manufacturer to adjust the ingredients to lower the COGS. So either i
140.
by m_ke 2y ago
We told them we didn't have time to work on something custom like that
141.
by m_ke 2y ago
I had one of the largest retailers to reach out to my company with an interesting problem. Given a weight and ingredients of packaged food they wanted to predict the content of each ingredient so that they could calculate the suppliers COGS
142.
by m_ke 2y ago
Literally the best government person we have had for competitive markets in decades
143.
by m_ke 2y ago
Please name some of the terrible parts EDIT: to make it easier here's a list of actions from perplexity: Here are more explicit actions taken by the FTC under Lina Khan's leadership: Lawsuit against Amazon (2023): The FTC filed a
144.
by m_ke 2y ago
1. If everyone is running custom written software that's not shared with anyone else if will be really hard to find vulnerabilities in them 2. I'm sure LLMs are already way better at detecting vulnerabilities than the average engi
145.
by m_ke 2y ago
Most people don’t want cloud hosted subscription software, we do it that way because VCs love vendor lock in and recurring revenue. Old school desktop software takes very little maintenance. Once you get rid of user tracking, AB testing, mo
146.
by m_ke 2y ago
Your customers don't care how your code smells, as long as it solves their problem and doesn't cost an arm and a leg. A ton of huge business full of Sr Principal Architect SCRUM masters are about to get disrupted by 80 line ChatGP
147.
by m_ke 2y ago
> Have you worked as a software engineer for a long time? I don't want to assume anything, but all of your points thus far read to me like they're coming from a place of not having worked in software much. Yes I've been a
148.
by m_ke 2y ago
That's not what I'm saying at all. I'm saying that there's a trend showing that you can improve LLM performance significantly by having it generate multiple responses until it produces one that meets some criteria. As an
149.
by m_ke 2y ago
There's a lot of work showing that we can reliably get to or above human level performance on tasks where it's easy to sample at scale and the solution is cheap to verify.
150.
by m_ke 2y ago
Sorry I wasn't clear enough, the cheap to validate part is only needed to train a large base model that can handle writing individual functions / fix bugs. Planning a whole project, breaking it down into steps and executing each o

More ›