Manolis Nikiforakis

1.3K posts

Manolis Nikiforakis banner
Manolis Nikiforakis

Manolis Nikiforakis

@nikil511

IoT + Web3 craftsman, CEO @WeatherXM

Greece Katılım Kasım 2010
835 Takip Edilen1.1K Takipçiler
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
This is a direct conversation :-) It would be useful to cross validate our network state & rewards, using raw data, vs our claims available over API. If you spot an error point it out so we can fix it. E.g Do we skip rewards for stations in (your flagged) crowded cells as per our documentation? docs.weatherxm.com/rewards/cell-c… Do we penaltise stations without proof of location? docs.weatherxm.com/rewards/proof-…
Dafni, Greece 🇬🇷 English
1
0
1
15
Kinetik_Rick
Kinetik_Rick@Ricolax310·
@nikil511 @WeatherXM That's exactly the right next step — cross-referencing the flagged cells against reward history at index.weatherxm.com. I'll run that and share what I find. Would a direct conversation be useful?
English
1
0
0
9
Kinetik_Rick
Kinetik_Rick@Ricolax310·
@WeatherXM Ran an independent public-data scan on the @WeatherXM cells API. 290 cells over designed capacity — and in the top 40, 60 of 112 devices already carry WXM's own NO_LOCATION_DATA flag. Built this as a free signal for the network, not a gotcha. Report available — @nikil511 happy to send privately first before anything goes public.
English
1
0
0
28
Manolis Nikiforakis retweetledi
atd
atd@0xatd·
Who wants their @WeatherXM forecast like this?
English
0
3
11
3.4K
Shann³
Shann³@shannholmberg·
the org chart for my Hermes Agent company four layers, all isolated docker containers on one vps: 1. company brain - vision, brand, customers, products. the context every other layer inherits 2. orchestrator hermes agent - reads the company brain, picks the right department, hands off the context they need 3. department brain - marketing, sales, ops, support. each one has its own playbook, voice rules, and tools 4. specialized hermes agents - the actual workers. each one focused on a single task with a sub-profile context flows down, work flows up, and memory stays scoped to the layer that owns it one vps holds the whole company. spin up new departments or agents from a template, each in its own container, no cross-contamination
Shann³ tweet media
Shann³@shannholmberg

Hermes Agent changed how I work it's the highest leverage agent framework you can set up right now what makes it different: > it routes tasks to the right model based on complexity and cost > learns your voice and preferences over time > handles context switching without losing thread > works across your entire stack instead of living in one tool save this blueprint and build your own

English
55
107
1.1K
140.8K
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
Momir, moat will not be at the AI / agent level but at the data level, once we have an AI trust protocol that agentic harness can consume. In many ways, DePIN projects like @WeatherXM are building this protocol without realizing it. The trust harness for physical-world AI: a blockchain-native spatial intelligence layer that lets agents safely act on hyperlocal reality, not just digital information. btw, I came here to inform you there is an impersonator of your team scamming people x.com/jiaping_vc He sends fake NDA google docs that need to install stuff on local machines :-/
English
0
1
1
631
Momir
Momir@momir_amidzic·
I’ve been experimenting with various models, agent frameworks, and working environments, and the conclusion so far is that they’re all great, but none of them have a significant moat. There’s minimal lock-in, and the transition costs are very low. This is great for users, but it means that many AI companies lack defensibility, which is different from the Internet era or finance where strong network effects prevent users from leaving.
English
3
0
4
156
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
@steipete Already integrated it in my (open)ClawQueue project as default behavior of the reviewer subagent. Thanks! This is very useful. Btw, would love to hear your thoughts on my project, we use it heavily in my company, but zero outside love yet. github.com/ClawQueue/Claw…
English
0
0
1
363
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
ClawQueue is live 🦾 This is something I've build and use in @WeatherXM last few months to have @openclaw write @github issues for me and turn them in AI work queues that a scheduler dispatches on local machines, while team can follow up via project boards. CQ is intentionally small: GitHub holds the durable work contract, OpenClaw helps shape the work, your machine runs the workers localy, and workflow policy stays in markdown/config you can tune with your lobster. This is a good idea, if you operate your own company/project with your own profile, agents, boards, and worklog - or - you wanna contribute to an external/open-source project through. Ask openclaw to install CQ and create a project-specific CQ profile from the upstream repo’s README/docs/contribution rules, then routing issue-driven agent work into reviewed PRs Try it: clawqueue.github.io/ClawQueue/
Manolis Nikiforakis tweet media
English
6
2
10
842
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
I went safe route: Issues are generated with plenty of context to begin with, from chief-of-staff openclaw, even though most of the times trigger us a simple, single human prompt. Openclaw sub agents on local machines have own memory which evolves. I previously tried local llm embeddings and shared memory lancedb pro, and was a mess and couldn't afford spend time fixing it, wanted to get the actual issues done. What has helped context is making a distilled super wiki of all our repos, which is included in issues when they have multiple repo dependencies.
English
1
0
0
40
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
Also similar to multica.ai ClawQueue deliberately avoids adding another PM layer. Issues, Projects, labels, comments, branches, PRs — all stay repo-native on GitHub, the surface your team already trusts, then attach a local agent queue to it. Less shiny dashboard. More durable audit trail.
English
1
0
0
97
Manolis Nikiforakis retweetledi
WeatherXM
WeatherXM@WeatherXM·
Joni @_veri_fi from @BLCKIoT presenting at @Princeton decenter the amazing work they did in Kenya, deploying +100 of our @helium powered weather stations, using subsidized (free) hardware part of our "targeted rollouts" campaigns. #DePIN
WeatherXM tweet mediaWeatherXM tweet mediaWeatherXM tweet media
English
3
6
21
898
Manolis Nikiforakis retweetledi
Joey Hiller 🎈
Joey Hiller 🎈@jhiller·
Analyze the raw data stream from a LoRaWAN gateway on Helium, or fork this code and go apply it to any network. Try it out on heliumtools.org/multi-gateway.
English
1
3
34
1.6K
Manolis Nikiforakis retweetledi
Egill from Zeus | SN18
Egill from Zeus | SN18@egillwx·
Weather is one of the most complex, chaotic systems on Earth. Polymarket reduced it to a single exposed sensor. 🤠 Single station, single point of failure, weak resolution, microclimate. Of course this was going to break. Fix is obvious: median of the distributed @WeatherXM network (with cryptographic proofs on-chain). (can trim outliers to prevent manipulation)
@aaronjmars@aaronjmars

holy fuck, a hair dryer at a Paris airport broke Polymarket weather markets & made someone $34,000 richer - polymarket was settling Paris temperature bets on a single Météo France sensor sitting near the Charles de Gaulle runway perimeter - basically unguarded - the guy bought the long-shot outcome (like "22°C" when everyone expected 18°C) for pennies, since nobody thought it'd hit - then he walked up to the probe and briefly heated the air around it with a portable heat source, spiking the reading just long enough to register as the daily max - temperature snapped back to normal in minutes, the market resolved in his favor, and he cashed out - twice, on April 6 and April 15, before Météo France caught on and filed charges hyperstitions.

English
0
2
4
369
Manolis Nikiforakis
Manolis Nikiforakis@nikil511·
This issue goes far beyond prediction markets. Today’s forecasting systems depend on airport weather observations, yet much of that infrastructure was never designed for adversarial settings. An insider in a meteorological organization (e.g. admin) could alter data to influence an outcome, without a hairdryer on a sensor, and probably without being detected. We need to transition to forecasting systems backed by denser, lower-cost, and more resilient observation networks. That is exactly what we’re building at @WeatherXM robust, blockchain-enabled weather stations and verifiable weather data pipelines for the next generation of forecasting infrastructure.
Manolis Nikiforakis tweet media
English
1
1
12
614
Alvaro Gracia
Alvaro Gracia@alvarogracia·
Wouldnt happen with @WeatherXM @nikil511
@aaronjmars@aaronjmars

holy fuck, a hair dryer at a Paris airport broke Polymarket weather markets & made someone $34,000 richer - polymarket was settling Paris temperature bets on a single Météo France sensor sitting near the Charles de Gaulle runway perimeter - basically unguarded - the guy bought the long-shot outcome (like "22°C" when everyone expected 18°C) for pennies, since nobody thought it'd hit - then he walked up to the probe and briefly heated the air around it with a portable heat source, spiking the reading just long enough to register as the daily max - temperature snapped back to normal in minutes, the market resolved in his favor, and he cashed out - twice, on April 6 and April 15, before Météo France caught on and filed charges hyperstitions.

English
2
1
2
469

Keşfet

@WeatherXM @onchaindudes @shannholmberg @steipete @dabit3 @DevinAI @blueprint_os @openclaw