lmeyerov

52 ms·

691.
by lmeyerov 4y ago
I tried with an oncologist who uses GPT ~daily. It frequently gives outdated answers that violate standard of care. No way to know which sentences are the wrong ones. The result is every fact must be searched again manually anyways. There i
692.
by lmeyerov 4y ago
Selling on revenue multiple is easier bc means leaving a lot of $ on the table: a good sale is in terms of the value to the acquirer, which can be a % of a massive #, not the immature revenue operations of some young startup . But if you d
693.
by lmeyerov 4y ago
The goal of lang features like these is to automate the optimization especially when it is hard to do at scale without it. Scenarios that start with "If only everyone..." are great candidates. Conversely, agreed, it's not THA
694.
by lmeyerov 4y ago
There is down and there is cratered 80% YoY: https://news.crunchbase.com/venture/global-vc-funding-monthl... So the question is what happens to VC-dependent companies - do exuberant times return and they continue thei
695.
by lmeyerov 4y ago
B sounds great but too much of the pydata ecosystem breaks here, like `class MyNeuralNet(nn.Module)`
696.
by lmeyerov 4y ago
Yes we are incorporating into Graphistry as part of how we help sec/fraud/misinfo/crime/etc analyst teams investigate their data. Our platform does all sorts of GPU visual graph analytics & graph AI once data gets lo
697.
by lmeyerov 4y ago
Afaict openai too, basically paying themselves for massive GPU bills
698.
by lmeyerov 4y ago
Loosely measure search, CVs, and job ads: https://gradientflow.com/the-vector-database-index/ This kind of analysis is rarely precise, but is useful for rougher tasks like tiering My personal question is if vector inde
699.
by lmeyerov 4y ago
Ah yes, thanks! Yeah we saw faiss + es leaders for serving embeddings / vector search, and pinecone / weaviate / I think milvus as next tier, so was curious if we could improve the analysis :)
700.
by lmeyerov 4y ago
I'm curious why you say 'more common'. When we did some adoption analysis, we found elastic vector indexes to have much more traction than others (excluding say faiss), even if it's a small % of elastic's advertisin
701.
by lmeyerov 4y ago
Neato, sounds like bitfusion in their early days! Definitely of interest to us, even w/ latency limits, both for ai dev & investigations and occasional full runs I do have to wonder about the non-oss licensing, as that's part
702.
by lmeyerov 4y ago
You should always be able to know revenue, so valuation is x 2 if a consultancy and x 10 if software. Likewise, increase/decrease based on whether revenue is doubling/tripling annually or not. The average multiple goes up and down
703.
by lmeyerov 4y ago
The interesting difference, and current problem, is VCs expect raised funds to be spent as burn on hiring more people despite existing people already being paid from VC capital instead of revenue. And they expect that to keep happening: eac
704.
by lmeyerov 4y ago
VC $ works out the same. 18-24mo, with an even higher burn than a bank would accept, and the assumption that you can raise a 2X+ bigger round at the end of it to pay for the previous staff + next round of hires.
705.
by lmeyerov 4y ago
Graphistry | Backend Engineer | Remote | US + global | Full-Time or Contract-to-Hire | S/TS clearance a plus if DC $100K-200K/yr + early-stage equity Graphistry is profitably growing 4x YoY as the first visual graph AI platform (
706.
by lmeyerov 4y ago
Ah yeah, I'd still fix the asymptotics first before doing multicore: 10X multicore speedup on a problem means little when haven't done the 100X+ speedup :) we do 500k+ node graphs on frontend, and million/billion on backend..
707.
by lmeyerov 4y ago
graph embedding is super useful for both viz + decision procedures - Viz: Embedding gives x/y coords, and if they rendered the edges, a traditional graph view. A cool thing about recent shift to embedding approaches to the graph drawin
708.
by lmeyerov 4y ago
I'm curious what makes it for embedding large graphs - afaict the core is a spring model implemented as O(n^2) in nodes and single-threaded? Wasn't sure of the _size param for the nested loop, maybe I'm misreading. Was hoping
709.
by lmeyerov 4y ago
I asked around similarly earlier this year. Some money goes somewhere... But likely not as much, and not widely distributed. In my convs with GPs, most LPs are very happy to avoid capital calls right now, as they went too far in during the
710.
by lmeyerov 4y ago
The funding market dropoff for Series A+ startups is a big deal for startup employees assuming they can keep high 6 figure salaries at vc-dependent companies where, in recent years, fundraising has far exceeded revenue. Ex: https:/&#x
711.
by lmeyerov 4y ago
I like the author started with measuring and thinking bandwidth, which makes sense for streaming through a big file, so I'd have continued that way towards a diff design & conclusion Continuing with standard python (pydata) and ok
712.
by lmeyerov 4y ago
As part of launching our global GPU edge network, we need to support low-volume regions, which means a small number of T4 gpu in different timezones. Azure ran out last Christmas, or at least refused us capacity, and is only adding the next
713.
by lmeyerov 4y ago
100% agreement, spot on observations! Indeed, we end up referring folks to cool diagramming tools all the time at Graphistry as d2/mermaid/diagrams.net are well-optimized for quick & beautiful diagram presentation tasks on man
714.
by lmeyerov 4y ago
Sorry for the Kite team, but for other folks aspiring here, more optimism is in order: We pay for copilot for Graphistry staff because it works well. Similar story for Docker Desktop: It's new, yet people are already paying for it to t
715.
by lmeyerov 4y ago
Having evaluated them -- Road 1: Sale to usual suspects like Palo Alto (though that window is closing due to raising $100M) or Cisco (that window may close if they raise again?). It is basically modern vpn, though will be years with a big e
716.
by lmeyerov 4y ago
This. Most ML people we work with, including new hires, are CPU-first and don't think in bandwidth/throughput, so we have to untrain basically every DS hire to this. Nvidia SW+HW is so far ahead here as they got it early. Sure, In
717.
by lmeyerov 4y ago
Good to see the pivot before cash ran out, but sad news :( Good luck to the team! IMO Mighty is too early bc they didn't tackle the SW side, 'just' the remote nature, so couldn't get their 10X. The 10-100X shift today co
718.
by lmeyerov 4y ago
And yet security VCs are investing here again and yeah, no idea why This Time Is Different We do get VDI users for our tool in some high-end security sensitive places to work around weak clients (budget is not uniformly distributed across u
719.
by lmeyerov 4y ago
Going public is a qualified event for a SAFE to convert -- it is a financing -- but curious about the dividends SAFEs are often capped, so taking awhile doesn't matter so much, except maybe interest calcs
720.
by lmeyerov 4y ago
Very much this. Same reason it's nice to use Stripe so you're not the one holding your customer's credit cards. Holding anyone's passport photo sounds horrible.

More ›