<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: mnbbrown</title><link>https://news.ycombinator.com/user?id=mnbbrown</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 11 Jun 2026 02:57:09 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=mnbbrown" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by mnbbrown in "PgDog is funded and coming to a database near you"]]></title><description><![CDATA[
<p>I've loved using pgdog for the last 6 months. It's been incredibly stable. It's nifty how they've solved the LISTEN/NOTIFY on a transaction pooler problem.</p>
]]></description><pubDate>Wed, 10 Jun 2026 15:09:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=48477525</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=48477525</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48477525</guid></item><item><title><![CDATA[New comment by mnbbrown in "Cohere Transcribe: Speech Recognition"]]></title><description><![CDATA[
<p>Added gladia..<p>- 1. Soniox (stt-async-v4): +176 new cases, running total 176/248 (71.0%)<p>- 2. ElevenLabs (scribe_v2): +26 new cases, running total 202/248 (81.5%)<p>- 3. Speechmatics (enhanced): +12 new cases, running total 214/248 (86.3%)<p>- 4. NVIDIA Parakeet (TDT 0.6B v2): +6 new cases, running total 220/248 (88.7%)<p>- 5. Mistral (voxtral-mini): +3 new cases, running total 223/248 (89.9%)<p>- 6. Gladia: +2 new cases, running total 225/248 (90.7%)<p>- 7. AssemblyAI (universal-2): +1 new cases, running total 226/248 (91.1%)<p>- 8. Deepgram (nova-3): +1 new cases, running total 227/248 (91.5%)<p>- 9. Cohere (transcribe-03-2026): +0 new cases, running total 227/248 (91.5%)<p>- 10. AssemblyAI (universal-3-pro): +0 new cases, running total 227/248 (91.5%)</p>
]]></description><pubDate>Wed, 01 Apr 2026 15:22:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=47602117</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=47602117</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47602117</guid></item><item><title><![CDATA[New comment by mnbbrown in "Cohere Transcribe: Speech Recognition"]]></title><description><![CDATA[
<p>They are all transcribed by multiple blinded "accent natives".
But yes, your point is valid - going to see if I can tease out the "single person accuracy".</p>
]]></description><pubDate>Wed, 01 Apr 2026 13:18:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=47600463</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=47600463</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47600463</guid></item><item><title><![CDATA[New comment by mnbbrown in "Cohere Transcribe: Speech Recognition"]]></title><description><![CDATA[
<p>Ran it over our internal dataset of ~250 recordings of people saying british postcodes (all kinds of accents, etc) - it's competitive for sure!<p>Soniox (stt-async-v4): 176/248 (71.0%)
ElevenLabs (scribe_v2): 170/248 (68.5%)
AssemblyAI (universal-3-pro): 166/248 (66.9%)
Deepgram (nova-3): 158/248 (63.7%) 
AssemblyAI (universal-2): 148/248 (59.7%) 
Cohere (transcribe-03-2026): 148/248 (59.7%) 
Speechmatics (enhanced): 134/248 (54.0%)<p>P.s. how do I get this to render correctly on here?</p>
]]></description><pubDate>Tue, 31 Mar 2026 20:43:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=47593226</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=47593226</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47593226</guid></item><item><title><![CDATA[New comment by mnbbrown in "Voxtral Transcribe 2"]]></title><description><![CDATA[
<p>Incroyable! Competitive (if not better) than deepgram nova-3, and much better than assembly and elevenlabs in basically all cases on our internal streaming benchmarking.<p>The dataset is ~100 8kHz call recordings with gnarly UK accents (which I consider to be the final boss of english language ASR). It seems like it's SOTA.<p>Where it does fall down seems to be the latency distribution but I'm testing against the API. Running it locally will no doubt improve that?</p>
]]></description><pubDate>Wed, 04 Feb 2026 22:03:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=46892471</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=46892471</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46892471</guid></item><item><title><![CDATA[GoCardless (YC S11) sells for €1.05B]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.ft.com/content/7904647c-91ea-4e90-b496-5d249a7a5d91">https://www.ft.com/content/7904647c-91ea-4e90-b496-5d249a7a5d91</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46231987">https://news.ycombinator.com/item?id=46231987</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 11 Dec 2025 14:44:06 +0000</pubDate><link>https://www.ft.com/content/7904647c-91ea-4e90-b496-5d249a7a5d91</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=46231987</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46231987</guid></item><item><title><![CDATA[Founder-led sales, from the perspective of a Founder, leading sales]]></title><description><![CDATA[
<p>Article URL: <a href="https://substack.com/inbox/post/180021176">https://substack.com/inbox/post/180021176</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46067161">https://news.ycombinator.com/item?id=46067161</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 27 Nov 2025 08:44:46 +0000</pubDate><link>https://substack.com/inbox/post/180021176</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=46067161</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46067161</guid></item><item><title><![CDATA[The Economics of Model Intelligence]]></title><description><![CDATA[
<p>Article URL: <a href="https://substack.com/home/post/p-176201646">https://substack.com/home/post/p-176201646</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45912495">https://news.ycombinator.com/item?id=45912495</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 13 Nov 2025 08:57:06 +0000</pubDate><link>https://substack.com/home/post/p-176201646</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=45912495</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45912495</guid></item><item><title><![CDATA[Using the node 18 native test runner with TypeScript and React]]></title><description><![CDATA[
<p>Article URL: <a href="https://matthewbrown.io/2025/09/04/node-test-runner">https://matthewbrown.io/2025/09/04/node-test-runner</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45125612">https://news.ycombinator.com/item?id=45125612</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 04 Sep 2025 10:21:42 +0000</pubDate><link>https://matthewbrown.io/2025/09/04/node-test-runner</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=45125612</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45125612</guid></item><item><title><![CDATA[New comment by mnbbrown in "Ask HN: Who is hiring? (September 2025)"]]></title><description><![CDATA[
<p>Elyos (<a href="https://elyos.ai">https://elyos.ai</a>) | London, UK | ONSITE | YC S23<p>Company: We're the UK's fastest growing AI voice company (~30% MoM) - very deeply focused on the trades industry right now (think your friendly neighbourhood plumbing company).<p>We're a small but experienced team hiring 3 more to join:<p>- Founding Engineer
 - Founding SDR
 - Founding Operations Lead (customer success)<p>Tech: GKE, python, postgres, telephony/media streaming, realtime LLMs, contextual eou detection, the occasional REST API.<p><a href="https://careers.elyos.ai/">https://careers.elyos.ai/</a> or message me - matt (at) elyos.ai<p>Mention bananas for extra kudos.</p>
]]></description><pubDate>Mon, 01 Sep 2025 20:35:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=45096398</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=45096398</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45096398</guid></item><item><title><![CDATA[20 days, one porsche, and a Greek wedding]]></title><description><![CDATA[
<p>Article URL: <a href="https://matthewbrown.io/2015/07/14/london-to-greece-porsche-997">https://matthewbrown.io/2015/07/14/london-to-greece-porsche-997</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=44279042">https://news.ycombinator.com/item?id=44279042</a></p>
<p>Points: 4</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 14 Jun 2025 21:43:16 +0000</pubDate><link>https://matthewbrown.io/2015/07/14/london-to-greece-porsche-997</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=44279042</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44279042</guid></item><item><title><![CDATA[New comment by mnbbrown in "Creating Bluey: Tales from the Art Director"]]></title><description><![CDATA[
<p>As someone who’s from Brisbane but spent the last 7 years in London you’re 100% correct. Brisbane is the best city in the world. I’m excited to eventually move back.</p>
]]></description><pubDate>Sat, 03 May 2025 08:45:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=43877702</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=43877702</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43877702</guid></item><item><title><![CDATA[Defining AI Tools in Python]]></title><description><![CDATA[
<p>Article URL: <a href="https://matthewbrown.io/2025/03/26/ai-tools-in-python">https://matthewbrown.io/2025/03/26/ai-tools-in-python</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43480745">https://news.ycombinator.com/item?id=43480745</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 26 Mar 2025 10:47:18 +0000</pubDate><link>https://matthewbrown.io/2025/03/26/ai-tools-in-python</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=43480745</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43480745</guid></item><item><title><![CDATA[Async Pytest in Python 3.11+]]></title><description><![CDATA[
<p>Article URL: <a href="https://matthewbrown.io/2024/05/19/notes-on-pytest">https://matthewbrown.io/2024/05/19/notes-on-pytest</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43146095">https://news.ycombinator.com/item?id=43146095</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Sun, 23 Feb 2025 03:15:12 +0000</pubDate><link>https://matthewbrown.io/2024/05/19/notes-on-pytest</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=43146095</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43146095</guid></item><item><title><![CDATA[Ask HN: iOS STT for postcodes, addresses better than deepgram?]]></title><description><![CDATA[
<p>Why is iOS speech to text so much better at numbers, addresses and postcodes than Deepgram, whisper, etc?<p>It nails them - the rest are mediocre at best. Does anybody know of a model that's accurate with this sort of data?</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=42914633">https://news.ycombinator.com/item?id=42914633</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 03 Feb 2025 03:32:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=42914633</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=42914633</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42914633</guid></item><item><title><![CDATA[New comment by mnbbrown in "Ask HN: What are you working on? (October 2024)"]]></title><description><![CDATA[
<p>That’s cool!<p>Very familiar with BMSs but the lack of open standards and protocols has been extremely frustrating - makes me appreciate how good we have it with HTTP, etc.<p>Lots say they support BACnet but that’s only if they’ve been configured and the points exported, etc.<p>Haystack is a great step forward for labelling too but adoption seems fill with complexity :)</p>
]]></description><pubDate>Mon, 28 Oct 2024 23:32:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=41977719</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=41977719</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41977719</guid></item><item><title><![CDATA[New comment by mnbbrown in "Ask HN: What Are You Working On? (October 2024)"]]></title><description><![CDATA[
<p>Decarbonising buildings and massive warehouses..<p>It's a very fun mix of hardware (for data collection), and crazy SQL queries to model energy flows between buildings, solar, batteries, etc. Considering just one building is pretty easy:<p>consumption = imported - exported + generated - stored + dispatched.
carbon = carbon intensity * imported
cost = tariff * imported<p>but then you add a site with a couple of buildings, solar on one of them, grid limited exports, etc modelling these flows is challenging. Like consider the case where one building got 10% of it's imported power from another building's excess solar, then calculating carbon becomes more difficult.<p>and once you've figured all that - then you have to figure out what makes commercial sense to do next.. install a battery, expand solar, move onto a TOU tariff, do nothing - and that's a whole other world of optimisation problems.</p>
]]></description><pubDate>Mon, 28 Oct 2024 13:02:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=41970646</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=41970646</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41970646</guid></item><item><title><![CDATA[Ask HN: Please stop hosting engineering blogs on medium]]></title><description><![CDATA[
<p>So much knowledge stuck behind a paywall.</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=40364195">https://news.ycombinator.com/item?id=40364195</a></p>
<p>Points: 6</p>
<p># Comments: 4</p>
]]></description><pubDate>Wed, 15 May 2024 07:52:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=40364195</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=40364195</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40364195</guid></item><item><title><![CDATA[New comment by mnbbrown in "Lowercase – A simple way to take and share notes"]]></title><description><![CDATA[
<p>Could ask for a username the first time they publish. Low friction</p>
]]></description><pubDate>Sat, 04 May 2024 12:42:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=40257227</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=40257227</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40257227</guid></item><item><title><![CDATA[RouterPi – A Compute Module 4 Router]]></title><description><![CDATA[
<p>Article URL: <a href="https://blog.zakkemble.net/routerpi-compute-module-4-router/">https://blog.zakkemble.net/routerpi-compute-module-4-router/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=39854346">https://news.ycombinator.com/item?id=39854346</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 28 Mar 2024 17:00:34 +0000</pubDate><link>https://blog.zakkemble.net/routerpi-compute-module-4-router/</link><dc:creator>mnbbrown</dc:creator><comments>https://news.ycombinator.com/item?id=39854346</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39854346</guid></item></channel></rss>