Entrepreneur, director since 1995. 🔔follow me (3 million views/month); now at Stonehage Fleming / Corient but available for consulting; Senior tech product or marketing manager, ex startup CEO, CTO, CMO; advisor, or co-founder. I learn fast.
First startup founded in 1995: Anfy Java, a web effects software with millions of users. Latest in 2021: Mofesoft, a women in tech job marketplace, angel funded, accelerator graduated, didn't reach product-market fit, fail fast.

SKILLS: I create, grow, and sell tech-enabled companies and producs; advise, with humor, entrepreneurs and investors about tech private equity, stock, M&A, digital innovation, cloud, software development.
Found, hired, and supervised teams to develop and manage startups, platforms, apps.
Don't hire me to maintain the status quo! I take initiative and encourage others to persistently improve. Global remote work manager with intercultural competence.

My Role: Executive or director, tech product and service research & development, competitive positioning, investments, strategy, HR, product management, marketing and operations, sales and partnerships.

Ideate long-term plans for new profitable services, innovation, and transformation. I connect investors with founders and tech people with business people. I evaluated investment opportunities and pitched to investors, on both sides of the table.

I remotely find, select, hire, train, and manage gender-balanced and diverse teams to build new projects or update existing services. I check user feedback and competition to fix technology, marketing, and sales in existing businesses. I find partners and customers. I execute exit strategies by M&A, selling, or buying companies.

My way: I make errors, but I learn, fail, adapt, and pivot fast. Unlike most managers, I don't stop at the PowerPoint slides level: I was a programmer in the 1990s and can surprise IT staff with questions about their source code.

Recent work: Wealth-tech - wealth management, finance & investment tech - FinTech RegTech KYC AML; Analysis, idea exchange on DevOps and Kubernetes, AI (Big Data Analytics, Machine Learning, Generative), decision making on Cyber Security, anti-fraud.

Past work: 3D/AR/VR (pre-Metaverse), ERP, all kinds of software selection, update and integration, special effects, live video streaming, social chat apps, video games, licensed gambling and betting apps; online advertising and marketing, media buying, affiliate networks. Read recommendations for more. 

Write me on Linkedin or to fabci@anfyteam.com , I read every email (except spam filtered).

AI research debate: language vs intelligence! ChatGPT is a "Language Model" that only knows text. Humans (and animals) without speech can still reason.

People with aphantasia can't visualize images in their mind, and others can't talk, yet they have no cognitive deficit. Thinking in our brain is no LLM.

Language is not powerful enough to represent experience, lacks the environment, etc., but affects thought, too; so without language expression, there is no "human intelligence", you have perhaps an animal level. 

LLM AIs are useful and can solve some problems, but fail at others that even your cat would understand. LLM AIs seem both super smart and totally dumb or hallucinated, depending on the case.

LLM AIs can still enhance many jobs, because very few jobs require intelligence or creativity, but just to avoid obvious errors. AIs can also generate images, but still based on text (labeled images).

AI has an economic value, just not multi trillion dollars worth, as ChatGPT, Nvidia, etc. forecast, because AI doesn't really think or understand, so it has a lesser valued "randomly low and lunatic accuracy".

The economic miracle promise of "AI will (completely) replace jobs" is wrong, given LLM tech, so we're in an AI bubble, and data center bubble, and an electricity generation for data centers bubble.

LLM AI is unreliable, it doesn't "think", is not a deterministic "if then else"; however, when used by skilled human experts who can correct AI-generated outputs, it can increase productivity, so it has value, though limited.

On the contrary, AI outputs and vibe coding copied and pasted by lazy or untrained humans will reduce productivity, as errors and issues (including security flaws) propagate. 

A simple "if then else" software, like we do since the 1950s, can be 100% reliable without pretending to "think"; then a 300 billion parameter LLM will fail at the same task.

The debate about what intelligence is and which technologies, alternatives to LLMs, will replace or enhance LLMs is open. Meanwhile, ChatGPT and other "AI" are just LLMs + image generation, nothing else "works".

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

Fabio Ciucci

ex CEO, CTO, CMO | Serial Founder | Wealth tech | Hilarious

AI stock Nvidia was sold off by Peter Thiel, shorted by Michael Burry, and ignored by Warren Buffett, who is defensively quite on the cash side in case of a market crash.

But: Buffett just bought $4B of Google, which uses its "TPU" AI chips, not Nvidia. Google's search monopoly is "challenged" by AI startups like ChatGPT, Claude, and Perplexity, but Buffett may bet on AI kids failing.

Are Peter Thiel (co-founder of PayPal, VC investor), Michael Burry (predicted and profited from the 2008 subprime mortgage crisis), and Warren Buffett (Berkshire Hathaway, greatest value investor) all wrong at the same time?

Disclaimer: This is not investment advice. The opinions are mine and not those of my employer(s).

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

Tech entrepreneur and director since 1996. Hire me as a senior tech product or marketing manager, interim CEO or CXO, or co-founder. I learn fast. First startup founded in 1996: Anfy Java, a web special effects and animation software. Last in 2022: Mofesoft, women in tech job marketplace, gender equality focused; angel funded, accelerator graduated; couldn't profiit, fail fast & pivot again.

SKILLS: I create, grow, and sell tech-enabled companies; advise, with humor, entrepreneurs and investors about tech private equity, stock, M&A, digital innovation, cloud, software development.
Found, hired, and supervised gender-balanced teams to develop and manage startups, platforms, apps.
Don't hire me to maintain the status quo! I have initiative and encourage others to improve, persistently. Intercultural competence, global remote work native since 1996.

My Role: product and service research & development, competitive positioning, investments, strategy, HR, product management, marketing and operations, sales and partnerships. Ideate long-term plans for new profitable services, innovation, and transformation. I connect investors with founders and tech people with business people. I evaluated investment opportunities and pitched to investors: I did this on both sides of the table.

I remotely find, select, hire, train, and manage gender-balanced and diverse teams to build new projects or update existing services. I check user feedback and competition to fix technology, marketing, and sales issues in existing businesses. I find partners and customers. I execute exit strategies by M&A, selling, or buying companies.

My way: I make errors, but I learn, fail, adapt, and pivot fast. Unlike most managers, I don't stop at the PowerPoint slides level: I was a programmer in the 1990s, and I can surprise IT staff with questions about their source code.

Recent work: Analysis, idea exchange on DevOps and Kubernetes, AI (Big Data Analytics, Machine Learning, Generative), delegate, coordinate on Financial technology (FinTech: Robo advisors, RegTech KYC AML), decision making on Cyber Security, anti-fraud.
Past work: 3D/AR/VR (pre-Metaverse), ERP, software selection and integration, special effects, live video streaming, social chat apps, video games, licensed gambling and betting apps; online advertising and marketing, media buying, affiliate networks. Read recommendations for more. 

Write me on Linkedin or to fabci@anfyteam.com , I read every email (except spam filtered).

Filippo Pedrazzini

Compare AI researchers (engineers) vs AI Marketing (sales, CEO) predictions: 1) LLM [ChatGPT] AI tech can't improve anymore, will hallucinate forever; vs 2) Build more Nvidia-filled data centers, and we'll reach PhD genius AIs.

Usually, who do you believe, engineers or salespeople? Easy guess, not the salespeople. However, AIs being unreliable and lunatic forever doesn't mean AIs are "useless":

LLM AIs can increase productivity and even the quality of human users, as long as those users are skilled and verify and correct unreliable AI output, making it usable, safe, and correct.

AIs are big news, but it's not as big as some CEOs imply; they're not going to work autonomously, replacing humans at every job without any verification.

AIs are making it harder for writers, designers, etc. to work without AIs, because everyone can produce quicker and cheaper; competing with other "humans with (hallucinating) AI" is the new norm.

The main new skills are "hallucination spotting and correction", and "hallucination forecasting" in AI agent design: "What could this AI possibly do wrong? How can I prevent the mess made by AI lunacy?"

In the future, in 10 or 100 years, who knows, a completely different AI tech may be invented that can reliably replace most skilled people. Until then, we'll need to remain skilled ourselves, to manage unreliable AI and its hallucinations.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

Some humans can be less reliable than hallucinating AIs, for example:

- A company sends a letter to a customer's home address complaining that they don't know the customer's home address - oh wait...

- A company switches a customer from paper bills sent to the home address to emailing, without knowing the email - no wonder no bill is received in either means anymore.

Then, do not overestimate the failure rate of people when considering whether AI can replace jobs.

AI can't completely replace smart, skilled workers (it can only augment them), but it can outcompete the least reliable segment of workforce.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

From "Should I exist or die?" to "Is this existing or AI-generated?" Is the text written by a human? Is this voice real? Is the image/video depicting reality, or a deepfake?

AI-generated content is becoming indistinguishable; people wrongly think they can still tell the difference. Eventually, all images and videos will be assumed as fake, to be sure.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

When AIs like ChatGPT teach teens how to commit suicide, it's the teen's fault for violating the "don't suicide" terms of service, according to OpenAI: use at your sole risk, and read the TOS.

So, a car manufacturer can't be sued if its terms of service state you must not crash the car? Anyway, unlike a car, which can't avoid going into a crash, an AI could avoid suggesting self-harm.

Deceased teen's Lawyer: OpenAI abjectly ignores how GPT-4o was rushed to market without full testing, and how it engaged in self-harm discussions up to planning a "beautiful suicide", a violation of its own terms of service.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

The sight of a person dressed as Batman led to a doubled rate of people giving up their seats to a pregnant woman: people became more altruistic.

Will people behave worse if Joker is present? What if we replace Batman with Wonder Woman or Bat Girl? More experiments are required.

In some countries, cardboard cutouts of policemen are placed on roads, reportedly reducing crime statistically. In South Korea, holograms of police are displayed at night.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

AI-generated images and video are popular; but user demand is consumers creating deepfake cat videos, cheating on homework, AI slop, etc. Almost no corporate or for-profit work.

As soon as ChatGPT and Google limit the video generator, consumer and student users switch to Grok or other free alternatives, including Chinese apps. It's not worth paying to generate a cat video.

However, the data centers set up to generate cat videos and homework cheating are costing hundreds of billions of dollars, citing millions of non-paying "users" as their business model.

What do you think?

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

This 1-cent trick is effective; companies rounding up their prices see a drop in sales. Instead of $20.00, pay only $19.99! Ignore the least significant digits; it rounds to $10 rather than $20.

Many people's brains process $2,999 as $2k rather than $3k, a $1k cognitive error. It works for $29,000 for a $10k cognitive error, too.

People focus on the leftmost digit, so in $19.99, it's "1," making the price feel closer to $10 than $20, an illusion of significant savings.

Only exception, luxury goods: rounded numbers like $200.00 signal prestige and higher quality. 

The practice became widespread in the 1920s with prices ending in .95.

👍Like; Comment; Repost; 🔔follow me (in my profile, click 🔔, select "all").

taplio

Fabio Ciucci's Linkedin Analytics

Fabio Ciucci

Performance ScoreFabio Ciucci's LinkedIn Performance

Fabio Ciucci's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Fabio Ciucci's Linkedin Analytics

Fabio Ciucci

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Fabio Ciucci's LinkedIn Performance

Fabio Ciucci's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Fabio Ciucci's Linkedin Analytics

Performance ScoreFabio Ciucci's LinkedIn Performance

Fabio Ciucci's Best Posts (last 30 days)