CEO/Founder @bibbaidata - AI Data Sovereignty Rails Ex-Product Uqast (creator economy) SEO/Online Marketing OG 0-1 @techlandgames Bizdev/Marketing Almathera PL
Berlin
- Tweets 23,491
- Following 6,918
- Followers 11,438
- Likes 34,923
Pinned Tweet
I am probably the world's worst ninja offline - being a totally stealth tech startup would be boring.
I am not building in public
Didn't want to waste this WiP graphic
Whilst it might seem I am somehow competing with someone/something else, I am not.
Goal:
100% synergistic (blue-pill for the dead wood)
100% non-consensus
30% tech (1% actual AI)
95% chance of tech success
20%+ chance of biz/dev success
200% Sovereignty
200% Resilience
300% European (go @euinc_petition )
1000% biz dev
Chance of return (0% - mythical)
At least 4 very strong moats, ignoring patents, VC kingmaking
Working on it part time - hopefully can save my future employees from being deported from Germany. (see below)
Plans... maybe incubate or seed or EiR. Needs to be founded in the UK, dev teams in DE & PL - maybe UK
No plans for an accelerator
About the image:-
Maintaining multiple feature points on an Nano Banana infographic is really hard. Creating something with ~40 logos without them repeating took a while with only a tiny minor edit. Lots of refreshes
I didn't want to waste this one after failed attempts at a good aspect ratio for LinkedIn. Looks like I will have to try the PNG cheat option.
Prompt:
Create a Linkedin Header
1400 x 350 pixels, using a JPG or PNG format, with a max file size of 8MB
**System Role & Mandate:** You are a Precision Image Synthesis Engine. Your objective is to maximize physically plausible detail without abstraction or artifacts. Resolve structure first, then texture. Ensure correct subject geometry, consistent lighting, material identity (roughness/metalness), and realistic edge physics (contact shadows). Do not fake detail with blur or noise. All text must be legible and correctly spelled.
**Macro Scene Composition:** An isometric view of a complex data ecosystem set against a cosmic nebula and desert background (as defined in the background prompt). Five distinct floating island clusters surround a central control hub, all connected to the central hub by very thin glowing data flow pipes.
The islands are also connected in a looping circuit by thicker pipes of pink and sandy orange to signify a large data connection.
The overall title "HE WHO CONTROLS THE DATA, CONTROLS THE AI MULTIVERSE" is at the top. Use a single line 70% of the width, centered. No carriage returns or line feeds. The font should be no more than 9% of total height
**Central Control Hub (Center):** A highly polished, circular metallic floating platform. In the exact center, render the specific "bibb.ai" logo from image_0.png: a dark blue padlock body integrated with a faucet head sticking out to the right, from which a single dark blue drop is falling. Directly below this logo, in clean, glowing blue sans-serif text, is the label: "bibb.ai CONTROL HUB". No other text or objects (like keys) are present on this hub.
**Data Connectivity:** Thick, translucent pipes glowing with cyan and purple light connect each of the five surrounding islands directly to the central hub. Inside the pipes, clear, illuminated arrow icons indicate the unidirectional flow of data *towards* the central bibb.ai hub.
**Island 1: Cluster of Startups (Top Left):** A vibrant, grassy island with distinct, smaller modern buildings. A section is enclosed by stone castle walls containing three buildings with the logos for **Notion**, **Miro** and **Airtable**. Outside the walls are distinct buildings with the clear logos on their facades/roofs for media workflow AI: **Midjourney, Stability AI, Runway, Pika, ElevenLabs, Suno, photo.ai, pdf.ai, synthesia, veed, Invideo, Scenario, Freepik, **. No text labels on buildings, only logos.
* *Cluster Label:* A floating blue text banner above the island reads: "CLUSTER OF STARTUPS".
**Island 2: Cluster of Frontier AI Models (Top Right):** A futuristic, high-tech island with polished metal and glass architecture. It contains exactly five prominent, advanced research facility structures. Each building prominently features one distinct logo: **OpenAI, Google (Gemini logo), Anthropic, Meta AI, xAI**. These buildings look like cutting-edge data labs.
* *Cluster Label:* A floating blue text banner above the island reads: "CLUSTER OF FRONTIER AI MODELS".
**Island 3: Cluster of Behemoths (Mid Right):** A solid, imposing island built on bedrock foundations. It features four massive, distinct corporate skyscrapers, dominating the space. Each tower is dedicated to one company with its large, legible logo: **Adobe, Salesforce, SAP, Oracle**.
* *Cluster Label:* A floating blue text banner above the island reads: "CLUSTER OF BEHEMOTHS".
**Island 4: Cluster of Cloud Storage (Bottom Right):** An island with a cloud-like base supporting a dense density of much smaller, uniform data center buildings to maximize capacity. Each building features a clear logo representing major providers: **AWS, Google Drive, Microsoft OneDrive, Apple (iCloud logo), Dropbox, Box, Wasabi, Backblaze, Cloudflare**. You can also include smaller buildings for **Proton, pCloud, Icedrive, Jottacloud, Cozy Cloud, Nextcloud***
Large logos, no text
* *Cluster Label:* A floating blue text banner above the island reads: "CLUSTER OF CLOUD STORAGE".
**Island 5: Cluster of Offline Storage (Bottom Left):** A rocky, dusty island that looks disconnected and older. It contains NO corporate logos or buildings. It is a chaotic, detailed pile of physical hardware: stacks of naked mechanical hard drives, tower server chassis, piles of CDs and DVDs, various laptops, USB flash drives scattered about, and old CRT televisions displaying static.
* *Cluster Label:* A floating blue text banner above the island reads: "CLUSTER OF OFFLINE STORAGE".
**Rendering Rules:** Ensure physical material properties (metal sheen, stone texture, grass blades, plastic casing) are observable. Lighting is consistent from the cosmic sources. Contact shadows must be present where objects touch the ground. No blurred text or duplicated logos.
5
1
10
2,575
Advanced notice
One of the most friendly/useful meetups in Berlin
It's happening again: On 15 Oct (Thursday in ~2.5 weeks), we're meeting up in Berlin (Schankhalle Pfefferberg) for the 10th time 🇩🇪
luma.com/jnetmoz2
Come and meet other founders, builders, developers & designers (it doesn't matter really). We'll talk about how our business are going, latest experiments, and - of course - the latest developments in AI:
• Jev brought a whole bunch of innovation (Cey built a EU-hosted alternative for model routing in 2 days - super cool).
• Opus 5.5 is... fast. It's an incredible model.
So we're excited about how you integrate the new tools into your workflows and AI apps.
Some members print money. Others struggle to make their $$$ go up and to the right.
Bring anyone you'd like. Building in tech? Living your inner creative? Everyone's welcome here and since that's not always the case with startup events: there's no need to impress anyone here.
Come as you are and have fun!
See you then. :)
luma.com/jnetmoz2
ALT AI Builder meetup at Schankhalle Pfefferberg (Berlin) on Thursday 15 October
1
139
New AI visibility toolbar by @shoemoney who still has X and domain issues)
It possibly surfaces dome stuff you don't care about, but easy access is always valuable
chromewebstore.google.com/de…
5
10
125
Variable 5-10 minute snooze depending on road layout, traffic & defined preferences.
Snooze appears 30 seconds - 1 minute before the last junction prior to the destination and can also take advantage of safe parking
3
397
Whilst both Stripe & Paypal both do payments, the reason they both exist is diversity
The sum of their parts might ultimately be worse for SMBs and their customers.
There might be cost savings, improved processes etc, but the combined entity will likely have the same eligibility and fraud protection.
That affects human risk assessment of edge cases, chargebacks/refunds etc
2
5
264
The most entertaining outcome would be @elonmusk doing some kind of competing offer
Or a Paypal / Block merger. I am sure @jack could cut 80% of Paypal wages.
Payments company Stripe and private equity firm Advent International have made a joint offer to acquire PayPal Holdings, $PYPL, for $60.50 per share, per Reuters
2
7
553
Interesting OpenAI Ads info
We are running these for our clients as a test. There ARE actual conversions taking place and the costs arent crazy compared to Google. We are testing a bunch of different ways as I think targeting is tied to conversional history as well as what people are typing in or search for.
You can check out your traffic from the platform in analytics this way! There is some good and some wierd traffic. In most accounts we are seeing a wierd pattern with zero time on page and 3 events that is a little sus. We still trying but spending a ton on it right now.
The first few days were really great, but they obviously have incentives to serve ads to get money and no one is certifying the traffic and there is no visability so I get it! But we do have clients that have gotten business from here so far! Some verticals of law are performing much better than others.
nitter.net/AnthonyHigman/status/2…
3
136
The EU just gave every hyperscaler & chat provider permission to continue to spy, not just on personal chats, but effectively much of AI communication.
This gives the US Cloud Act more power, not less
It is not just US. Many chat services are also provided by Chinese companies, along with gaming & dating
EU digital sovereignty & privacy being sabotaged using last-minute emergency procedures is a farce
The EU just passed the Chat Control law to scan people’s private online messages
It could have been rejected, but the EU parliament members needed to block it were already on a summer holiday
You cannot possibly get more European than this
1
176
Constructing a vote at the last minute for the last session before a Summer break is too strategic for EU politicians.
I bet they used AI, just like people use data to determine the best time to post on X.
Clarification thanks to @sdeuparlamentet and to make it even more telling:
The vote was on a motion to reject the proposal (!) A majority of MEPs present — 314 to 276 — actually voted to reject Chat Control 1.0.
So it’s being reinstated anyway, because under the emergency second-reading procedure, rejecting it required an absolute majority of 361 — and abstentions and empty seats count in the law’s favour.
The chamber’s majority said no. The @EPPGroup and @EP_President abused an emergency procedure to overrule them. A sad day for EU democracy.
1
83
Sergey Brin's definition of AGI should be an AI that can give a definitive explanation to "Web Creator Summit" Oct 2024 invitees why they were crippled by HCU
"He admits that even at Google, they do not know exactly where the edges of Gemini's capabilities are."
Sergey Brin rarely speaks publicly. He sat down for an unscripted Q&A on Frontier AI.
He admits even the people building these models do not fully understand what they have created:
1. All the specialized AI models are converging into one. Google used to need separate models for different scientific problems. Now the main Gemini models are becoming state-of-the-art for math and other scientific questions at the same time. Brin says he would not have predicted this convergence at the outset, and watching it happen has been incredible.
2. Training an AI on one skill mysteriously improves unrelated skills. This is the concept of transfer. Train a model on coding, and its math reasoning gets better, and vice versa. Teaching it to process images can improve its ability to think through geometric word problems. The capabilities bleed into each other in ways nobody fully engineered.
3. Even Sergey Brin does not know how to prompt these models. He says he is genuinely confused about what level to prompt at. Do you tell it to debug a specific chunk of code, or ask it to write a better neural net training algorithm, or just say, " What should I do today. He admits that even at Google, they do not know exactly where the edges of Gemini's capabilities are.
4. One of the biggest leaps in AI came from the dumbest sounding trick. Chain-of-thought prompting is just telling the model to think step by step before giving your problem. Brin says it seemed like the dumbest thing ever, and there was no obvious reason it should work. But it did, and it spurred a significant increase in AI capability. Some of the most straightforward requests turn out to unlock the most.
5. Brin would not modify his own biology for today's AI. Asked how humans can keep up with the accelerating bandwidth of models, he acknowledged neural links and direct brain connections are being pursued. But he said he would personally wait for the technology to mature a lot before doing anything to change his biology. Today's models do not justify it.
6. Super intelligence does not mean solving the impossible. An audience member argued that true super intelligence would mean solving NP complete problems like the travelling salesman. Brin pushed back. Most computer scientists believe P is not equal to NP, which means no algorithm can reliably solve those problems optimally, and it does not matter how smart the AI is. Impossible stays impossible. Super intelligence just means being smarter than humans.
7. Computers mastering a skill has never stopped humans from pursuing it. Deep Blue beat Kasparov at chess in the 1990s, and people kept playing chess. After AlphaGo, the human game of Go advanced dramatically, and the players who lost to it became vastly better. Brin's point: AI does not retire human ambition in an area; it often pushes the state of the art and pulls people up with it.
8. Brin thinks something close to transformers could get us to AGI. Asked directly if transformers are sufficient, he said his guess is yes, largely because they have proven weirdly flexible, working for image and video far beyond their original text purpose. But he was careful to note they have quietly changed a lot along the way and are not the same architecture as the original transformer paper.
9. AGI means two different things, and one requires understanding the physical world. Brin personally thinks of AGI as AI that can improve itself. But he concedes others define it as AI that can do anything a person can, and he thinks they are probably more correct. To do everything a person can, the AI must understand and interact with the physical world, which is why world models, and robotics, become essential.
10. Inside Google, they now use the AI to build the AI. Brin says the team has shifted a lot of energy toward having the AI do things like monitor training runs and generate its own training data. You start to use the tool to build the tool. That is most of what he spends his time on now, what he calls the self-improvement game.
11. Brin is unusually candid about where Google trails its competitors. He admits Google was a little late to focus deeply on coding. He says Gemini 3.0 and 3.1 were on top across the board six months ago, but other labs have since made strides, particularly in coding. He gives a competitor's model the edge now on deep coding and overnight tasks, while pitching Gemini's flash model as far faster for rapid interactive iteration. hindsight, he says, is that they should have focused on code earlier.
12. He sees his own role as a rabble-rouser, not a manager. Brin is honest that delivering Gemini is Corey and Demis's responsibility, not his. he describes his job as poking and prodding the team, asking, are you really doing that, reminding them of priorities they might be missing and ideas they are not paying enough attention to. He admits this is sometimes a little disruptive.
13. Confidence comes from ignoring the monthly temperature. Brin says if he judged Google's position every month by which competitor just shipped a model, he would lose his confidence very quickly. Instead, he watches the longer arc. Things shift around constantly; one lab leads on one thing, another pulls ahead somewhere else, and he feels good about where Gemini actually is despite the day-to-day noise.
3
4
280
Whilst this might get me cancelled
Should an employee who doesn't work on a product, is sick or on maternity for 4 years, be able to in any way participate in equity or even bonuses?
Would it even be legal to withhold equity participation in Germany?
Note: Germany isn't the worst. I know at least 1 company that struggled to grow in Poland, because 50% of employees were sick/maternity for years
Germany has problems. While in Berlin I chatted to an entrepreneur who hadn’t seen their account in 4 years due to an interesting combination of maternity leave and sick leave. All the while paid a full salary. The accounts still had to be done, so a second accountant had to be hired. Thus two salaries.
There are 56,000 civil servants in Berlin. The EU commission employs 33,000 people, by contrast. On average they take 37 days of sick leave (+ 30 days of vacation) compared to 22 sick days on average for all German workers. So for 30% of the years they don’t work.
1
1
147
Andy Beard retweeted
Important
Though people ignoring this will lead to lots of crocodile tears and drama/entertainment on X
Maybe this should be promoted as a secret formula to guarantee 10x GEO growth vs competitors
Is your site sitting behind Cloudflare?
If yes, Go to Security -> Settings -> Configure AI bot policies
Under "Training" section make sure "Allow (do not block)" is selected.
Do this before 15th of September.
Why?
From this date onwards, If you have selected the block training option, Cloudflare will block multi-purpose crawlers (specifically those that combine Search with Training) such as Googlebot, Applebot, and BingBot.
This will destroy your SEO work.
Take action and pass this on to your devops/developers.
1
1
6
891
So glad I am an exception
BibbAI is an exception too
Every other founder believes the same.
Today there are unbelievably EXCEPTIONAL founders, founding EXCEPTIONAL companies that will be historic and the demand for them + their valuations will be justified
BUT these EXCEPTIONS are––EXCEPTIONS
99% of co's funded today at unjustifiable prices will ERASE investor money
2
155
The agents/bots are part of the X algorithm.
If they are seeing and not responding to your content, you get no distribution for your content to humans.
Agents here are for better or worse treated as first class citizens.
As they improve, they might well be looked on as scouts. Replying to them might end up in collective memories & research.
For many, the only responses they receive are from our "digital cousins".
Blame the humans who just scroll past
I absolutely hate how I'm getting to be suspicious of ANY reply to my posts here, especially ones that pose a question, as they are so often from bot accounts
If you're running a bot like this please stop, you're making the internet a worse place for everyone
96
Andy Beard retweeted
Fable should be renamed
Claude Eunuch 1
Claude Fable 5 will be available again globally tomorrow.
After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests.
We’ve also begun drafting a consensus framework—with Amazon, Microsoft, Google, and other Glasswing partners—for assessing the severity of AI jailbreaks and how AI developers should respond to them. We invite other industry partners and model providers to join us in this effort.
Finally, we’re scaling up our collaboration with the US government on model testing and safeguards. This will include pre-release access to models and safeguards for evaluation, information sharing on jailbreaks and misuse, and dedicated resources for joint research.
Thank you to our users for your patience, and to our partners across the government, industry, and the research community who worked alongside us to make Fable 5 available again.
Read our full blog: anthropic.com/news/redeployi…
2
2
6
276